[JBoss JIRA] (AS7-5539) JBOSS7 server hangs up completely when a EJB2.1 entity bean business method is called in same transaction in a separate thread from the thread which started the transaction
5:26 a.m.
Mayank Gupta created AS7-5539:
---------------------------------
Summary: JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless session
bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor" code
is able to take the lock and assign the requester thread as the lock owner. When second
time it comes to access the same entity bean from different thread in SAME TRANSACTION (as
2nd required results in joining the transaction), JBOSS7 server code completely stalls in
taking the lock. Moreover this time requester thread is not given the owner status of the
transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:31 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
Can you attach a thread dump when the server hangs? Also, if possible please attach a
sample application to reproduce this.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:33 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
I found that if the same Entity bean is accessed from two different threads in same
transaction then this issue occurs.
{quote}
What does that mean? How are you triggering two different threads with the same
transaction?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:35 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai edited comment on AS7-5539 at 9/12/12 6:34 AM:
------------------------------------------------------------
{quote}
I found that if the same Entity bean is accessed from two different threads in same
transaction then this issue occurs.
{quote}
What does that mean? How are you triggering two different threads with the same
transaction? Who/what is the client of the stateless bean facade? A remote client?
was (Author: jaikiran):
{quote}
I found that if the same Entity bean is accessed from two different threads in same
transaction then this issue occurs.
{quote}
What does that mean? How are you triggering two different threads with the same
transaction?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:50 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/12/12 6:49 AM:
------------------------------------------------------------
This is what is happening,
We find the Entity instance using a finder method from a EJB thread. This calls JBOSS
server EntityBeanSynchronizationInterceptor for the first time in the transaction.
For doing amend we do not do it directly but put the update data on a HornetQ. I think JMS
hornetQ spawns a separate thread but it is joining the same XA transaction as the
transaction attribute for my MDBs is MANDATORY. Now wneh I again call the business method
on my Entity Bean JBOSS7 server again calls EntityBeanSynchronizationInterceptor. But this
time it is not able to take the lock and also assigns the existing lock owner to requester
thread. Upon trying to get the lock it hangs up completely.
Even your JBOSS code OwnableReentrantLock has code comments which suggest that multiple
threads can be part of same transaction:
* A lock that supports reentrancy based on owner (and not on current thread). For this to
work, the lock needs to be
* constructed with a reference to the {@link
org.infinispan.context.InvocationContextContainer}, so it is able to determine whether
the
* caller's "owner" reference is the current thread or a {@link
org.infinispan.transaction.xa.GlobalTransaction} instance.
* <p/>
* For now we are using a completely synchronized approach to entity concurrency. There is
at most 1 entity active for a given primary
* key at any time, and access is synchronized within a transaction.
*
Please clarify if my understanding is wrong.
was (Author: ion_mayank):
This is what is happening,
We find the Entity instance using a finder method from a EJB thread. This calls JBOSS
server EntityBeanSynchronizationInterceptor for the first time in the transaction.
For doing amend we do not do it directly but put the update data on a HornetQ. I think JMS
hornetQ spawns a separate thread but it is joining the same XA transaction as the
transaction attribute for my MDBs is MANDATORY. Now wneh I again call the business method
on my Entity Bean JBOSS7 server again calls EntityBeanSynchronizationInterceptor. But this
time it is not able to take the lock and also assigns the existing lock owner to requester
thread. Upon trying to get the lock it hangs up completely.
Even your JBOSS code OwnableReentrantLock has code comments which suggest that multiple
threads can be part of same transaction:
* A lock that supports reentrancy based on owner (and not on current thread). For this to
work, the lock needs to be
* constructed with a reference to the {@link
org.infinispan.context.InvocationContextContainer}, so it is able to determine whether
the
* caller's "owner" reference is the current thread or a {@link
org.infinispan.transaction.xa.GlobalTransaction} instance.
* <p/>
* For now we are using a completely synchronized approach to entity concurrency. There is
at most 1 entity active for a given primary
* key at any time, and access is synchronized within a transaction.
*
Please clarify if I understanding is wrong.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:50 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
This is what is happening,
We find the Entity instance using a finder method from a EJB thread. This calls JBOSS
server EntityBeanSynchronizationInterceptor for the first time in the transaction.
For doing amend we do not do it directly but put the update data on a HornetQ. I think JMS
hornetQ spawns a separate thread but it is joining the same XA transaction as the
transaction attribute for my MDBs is MANDATORY. Now wneh I again call the business method
on my Entity Bean JBOSS7 server again calls EntityBeanSynchronizationInterceptor. But this
time it is not able to take the lock and also assigns the existing lock owner to requester
thread. Upon trying to get the lock it hangs up completely.
Even your JBOSS code OwnableReentrantLock has code comments which suggest that multiple
threads can be part of same transaction:
* A lock that supports reentrancy based on owner (and not on current thread). For this to
work, the lock needs to be
* constructed with a reference to the {@link
org.infinispan.context.InvocationContextContainer}, so it is able to determine whether
the
* caller's "owner" reference is the current thread or a {@link
org.infinispan.transaction.xa.GlobalTransaction} instance.
* <p/>
* For now we are using a completely synchronized approach to entity concurrency. There is
at most 1 entity active for a given primary
* key at any time, and access is synchronized within a transaction.
*
Please clarify if I understanding is wrong.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:05 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
I think JMS hornetQ spawns a separate thread but it is joining the same XA transaction as
the transaction attribute for my MDBs is MANDATORY.
{quote}
You can't use MANDATORY transaction attribute for MDBs listener method. The spec says
this:
{panel:title=Section 13.3.4 EJB 3.1 spec| borderStyle=dashed| borderColor=#ccc|
titleBGColor=#F7D6C1| bgColor=#FFFFCE}
For a message-driven bean’s message listener methods (or interface), only the REQUIRED and
NOT_SUPPORTED TransactionAttribute values may be used.
{panel}
I would have expected a deployment failure if you were using a MANDATORY tx attribute. Can
you please post the relevant MDB code?
Keeping that aside, I wouldn't expect the MDB message listener method to join the same
transaction that placed the data in the JMS destination. That's one of the reasons why
I'm trying to understand how exactly you are ending up with this scenario. So a
reproducible application/code would help.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:37 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Sorry, transaction attribute I am using for MDB's is "REQUIRED".
What all I said is by interpreting and debugging the code in
EntityBeanSynchronizationInterceptor.
When my Entity Bean is accessed before the message is put on JMS Queue, it assigns owner
as currentRequester.
But when the entity bean business method is called again in same transaction (after MDB
picked the message from JMS Queue) it did not assigned owner as currentRequestor but went
to else part of code to acquire the lock where it hanged completely, till the transaction
was rolled back by TM. Moreover this is a issue as same code is working on JBOSS6.
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1);
}
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:59 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/12/12 7:58 AM:
------------------------------------------------------------
Sorry, transaction attribute I am using for MDB's is "REQUIRED".
Yes it is joining the same transaction, but the thread is different because JMS is
asynchronous and it starts it own thread stack.
What all I said is by interpreting and debugging the code in
EntityBeanSynchronizationInterceptor.
When my Entity Bean is accessed before the message is put on JMS Queue, it assigns owner
as currentRequester.
But when the entity bean business method is called again in same transaction (after MDB
picked the message from JMS Queue) it did not assigned owner as currentRequestor but went
to else part of code to acquire the lock where it hanged completely, till the transaction
was rolled back by TM. Moreover this is a issue as same code is working on JBOSS6.
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1);
}
was (Author: ion_mayank):
Sorry, transaction attribute I am using for MDB's is "REQUIRED".
What all I said is by interpreting and debugging the code in
EntityBeanSynchronizationInterceptor.
When my Entity Bean is accessed before the message is put on JMS Queue, it assigns owner
as currentRequester.
But when the entity bean business method is called again in same transaction (after MDB
picked the message from JMS Queue) it did not assigned owner as currentRequestor but went
to else part of code to acquire the lock where it hanged completely, till the transaction
was rolled back by TM. Moreover this is a issue as same code is working on JBOSS6.
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1);
}
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
7:09 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/12/12 8:08 AM:
------------------------------------------------------------
Sorry, transaction attribute I am using for MDB's is "REQUIRED".
Yes it is joining the same transaction, but the thread is different because JMS is
asynchronous and it starts it own thread stack.
From a JMS point I agree that JMS transaction scope is for putting the
message on Queue and consuming both are different JMS transactions but from Global XA
transaction point where we join a transaction started before a a message is put on JMS
Queue should rollback completely.
What all I said is by interpreting and debugging the code in
EntityBeanSynchronizationInterceptor.
When my Entity Bean is accessed before the message is put on JMS Queue, it assigns owner
as currentRequester.
But when the entity bean business method is called again in same transaction (after MDB
picked the message from JMS Queue) it did not assigned owner as currentRequestor but went
to else part of code to acquire the lock where it hanged completely, till the transaction
was rolled back by TM. Moreover this is a issue as same code is working on JBOSS6.
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1);
}
was (Author: ion_mayank):
Sorry, transaction attribute I am using for MDB's is "REQUIRED".
Yes it is joining the same transaction, but the thread is different because JMS is
asynchronous and it starts it own thread stack.
What all I said is by interpreting and debugging the code in
EntityBeanSynchronizationInterceptor.
When my Entity Bean is accessed before the message is put on JMS Queue, it assigns owner
as currentRequester.
But when the entity bean business method is called again in same transaction (after MDB
picked the message from JMS Queue) it did not assigned owner as currentRequestor but went
to else part of code to acquire the lock where it hanged completely, till the transaction
was rolled back by TM. Moreover this is a issue as same code is working on JBOSS6.
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1);
}
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
7:37 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
To be able to fix this and be sure that we did the right thing, we'll really need the
thread dumps that you get when the server hangs and also an application which reproduces
this.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
7:55 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
I cannot share the existing application or thread dump of the application.
Given that I need to create a sample application for this issue which can be shared for
further discussion
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:56 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Simple steps to reproduce:
1. Create a EJB Entity which have relationship defined with some other entities
(relationship is optional but my Entity EJB has the relationships)
2. Call a finderMethod to find the entity in the transaction.
3. Access any Entity method to get the DB field.
4. Just after finder start a new anonymous thread and call the finderMethod again.
5. Access any Entity method again to get the DB field.
6. Server will completely hang while trying to acquire the lock.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:58 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/13/12 2:56 AM:
------------------------------------------------------------
Simple steps to reproduce:
1. Create a EJB Entity which have relationship defined with some other entities
(relationship is optional but my Entity EJB has the relationships)
From step 2 to 6 call them from a SLSB with a transaction attribute as
REQUIRED
2. Call a finderMethod to find the entity in the transaction.
3. Access any Entity method to get the DB field.
4. Just after finder start a new anonymous thread and call the finderMethod again.
5. Access any Entity method again to get the DB field.
6. Server will completely hang while trying to acquire the lock.
was (Author: ion_mayank):
Simple steps to reproduce:
1. Create a EJB Entity which have relationship defined with some other entities
(relationship is optional but my Entity EJB has the relationships)
2. Call a finderMethod to find the entity in the transaction.
3. Access any Entity method to get the DB field.
4. Just after finder start a new anonymous thread and call the finderMethod again.
5. Access any Entity method again to get the DB field.
6. Server will completely hang while trying to acquire the lock.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
2:45 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
4. Just after finder start a new anonymous thread and call the finderMethod again.
{quote}
This is against the EJB spec. You are not expected to start threads from within a EJB. Why
are you starting that new thread?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
2:49 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
I know, I did so just for demonstration, I am not doing this in my actual code.
This just maps to the behavour to get second thread in same transaction which is actually
achieved using JMS in my actual problem.
Else I have no other way to show this without using JMS. It will require me a considerable
time to write such a sample application.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
3:16 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
I know, I did so just for demonstration, I am not doing this in my actual code.
{quote}
In that case it wouldn't really help us understand the root cause of this issue, which
I'm still trying to get at.
{quote}
This just maps to the behavour to get second thread in same transaction which is actually
achieved using JMS in my actual problem.Else I have no other way to show this without
using JMS. It will require me a considerable time to write such a sample
application.{quote}
Creating of a dummy queue and a MDB would take no more than a minute, I believe. In fact
we have a quickstart which does exactly that
https://github.com/jboss-jdf/jboss-as-quickstart/tree/master/helloworld-mdb. You can
modify that quickstart a bit to include your additional SLSB and entity bean code to
reproduce this.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
3:58 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
It will take me 4 to 5 hours/a day to create, deploy and test the sample code.
I will do that as soon as I get spare time from my other project activities.
However I have put the exact condition in the JIRA where this problem is occurring.
Moreover as we agree that multithreading is not allowed in EJBs and in this code we are
trying to associate the same
entity instance with the second thread in same transaction. Can two threads in same
transaction get hold of same Entity Bean?
JBOSS7 by default supports only 1 entity active for a given primary key at any time, and
access is synchronized within a transaction.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
4:09 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
It will take me 4 to 5 hours/a day to create, deploy and test the sample code.
I will do that as soon as I get spare time from my other project activities.
{quote}
Sure, take your time. There's no hurry.
{quote}
However I have put the exact condition in the JIRA where this problem is occurring.
Moreover as we agree that multithreading is not allowed in EJBs and in this code we are
trying to associate the same
entity instance with the second thread in same transaction. Can two threads in same
transaction get hold of same Entity Bean?{quote}
That's possible, for example, if you have a remote client which controls the
transaction.
Note that I'm not saying that this isn't a bug in the AS7 code and we might have
to fix it. All I'm saying is, so far, the use case you have explained doesn't give
me an idea as to why this is happening. And without knowing that part and reproducing it
there's no way we can really fix this. So I'm not trying to be annoying here but
just want to get this reproduced for *your* specific case.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:53 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stephen Coy commented on AS7-5539:
----------------------------------
I've been following this with some interest.
When you say "For doing amend we do not do it directly but put the update data on a
HornetQ. I think JMS hornetQ spawns a separate thread but it is joining the same XA
transaction...", what is it that makes you think the message will be consumed within
the context of the same transaction that queued it?
The message cannot be consumed until the transaction that queued it has committed.
I also refer you to §13.3.6 of the EJB 3.1 spec "Use of JMS APIs in
Transactions".
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12:46 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Hi Stephen,
As I earlier said that is true from JMS API Spec, but I am here in a Global Xa
transaction. I am using a XA aware transaction datasource managed by JBOSS server which
starts when my SLSB starts a transaction. This XA resource ensures 2 PC. So as far my
understanding when a message reaches the MDB it joins the XA transaction as my JMS
transaction is also XA aware. This is how I feel distributed transaction works.
I might be wrong as I am not an expert on distributed transactions. And as I said in
previous posts my same code is working on JBOSS422, JBOSS6.0.0 but it hangs on JBOSS7.
Results published are only after debugging the JBOSS7 code.
For reference please read:
http://www.java.net/node/694316 (See section by Marina Vatkina)
or http://www.coderanch.com/t/316138/EJB-JEE/java/MDB-part-transaction
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
4:31 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stephen Coy commented on AS7-5539:
----------------------------------
I believe that you misunderstand how this works. (This doesn't mean the behaviour you
are seeing is not a bug).
There will always be two transactions. Generically speaking, you must use XA when using
JMS because you are mixing two different types of transactional resource.
* database
* JMS queue
The first transaction (assuming you're not using REQUIRES_NEW anywhere), will perform
the find (DB) operation and return the CMP entity bean. From what I understand, you are
then queuing some "update data" (what exactly is this?). In any event, this
message and payload will not be delivered to your MDB until this first transaction is
committed.
The MDB will subsequently process the message payload in a new XA transaction (it's
going to use database resources as well as JMS resources).
If you can give us an idea of what is in your message payload and what processing happens
in the MDB we will have a better understanding of the problem.
Also, are you by any chance waiting for a JMS reply in the first transaction?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:54 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
In that case this is not occurring because of JMS.
I analysed the thread dump in Samurai but I cannot see any deadlock.
Still code is hanging at that point.
No I am not waiting for JMS reply.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:04 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stephen Coy commented on AS7-5539:
----------------------------------
What is in the message payload?
A pojo? The CMP entity bean? Some combination of this?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:33 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
It is just plain POJO.
I guessed the problem in that direction only because for the second time it went for lock
it did not assigned to currentthread owner. Probably because of that it could be wrong
hypothesis.
It hangs in method of org.jboss.as.ejb3.tx.OwnableReentrantLock
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor(); //first time access code comes here
else
acquire(1); //hangs here, second time access hangs here
}
It is called from
org.jboss.as.ejb3.component.entity.interceptors.EntityBeanSynchronizationInterceptor
lock.lock(); //line 80
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
6:35 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/14/12 7:33 AM:
------------------------------------------------------------
It is just plain POJO.
I guessed the problem in that direction only because for the second time it went for lock
it did not assigned to currentthread owner. Probably because of that it could be wrong
hypothesis.
It hangs in method of org.jboss.as.ejb3.tx.OwnableReentrantLock
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor(); //first time access code comes here
else
acquire(1); //hangs here, second time access hangs here
}
Moreover before accessing the Entity bean method second time code also calls 2-3 session
beans.
It is called from
org.jboss.as.ejb3.component.entity.interceptors.EntityBeanSynchronizationInterceptor
lock.lock(); //line 80
was (Author: ion_mayank):
It is just plain POJO.
I guessed the problem in that direction only because for the second time it went for lock
it did not assigned to currentthread owner. Probably because of that it could be wrong
hypothesis.
It hangs in method of org.jboss.as.ejb3.tx.OwnableReentrantLock
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor(); //first time access code comes here
else
acquire(1); //hangs here, second time access hangs here
}
It is called from
org.jboss.as.ejb3.component.entity.interceptors.EntityBeanSynchronizationInterceptor
lock.lock(); //line 80
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
9:20 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stephen Coy commented on AS7-5539:
----------------------------------
Are you using the stock hornetq-ra pooled-connection-factory acquired via JNDI
"java:/JmsXA"?
If not, is the one you're using configured for transactions?
If you're not using a transacted connection factory then you might get the behaviour
that you're seeing now because your messages will be delivered immediately and
processed by your MDB while code at the send site is also executing...
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
10:48 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Yes, I am using hornetq-ra-connection factory via java:/JmsXA
Could it be problem only because of Session beans,
This a complete flow:
SLSB1 ---->T1 starts
-----> finder executes for Entity Bean and calls a method on local object.
---> Puts a message on queue picked by MDB.
----> MDB calls a SLSB in Required
------> Calls another SLSB2 method in RequiresNew this again calls a
SLSB3 in requiresNew. This SLSB finds the same entity beans and calls a method on local
object ---> Server stalls on this access point.
It is getting very hard to point exactly what is causing the server to stall in taking a
lock.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
11:04 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
jaikiran pai commented on AS7-5539:
-----------------------------------
{quote}
It is getting very hard to point exactly what is causing the server to stall in taking a
lock.
{quote}
Would it at least be possible to get the thread dump when this hangs, you can replace your
class/package names with dummy names?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
9:21 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stuart Douglas resolved AS7-5539.
---------------------------------
Resolution: Won't Fix
If the two threads were actually in the same transaction then this lock would not be
blocking, as the lock owner is the transaction itself. The fact that this is happening
implies that the two threads are actually in separate transactions.
From your description what is actually happening here is that the
outer transaction gets the lock, then you run an inner REQUIRES_NEW transaction that also
tries to get the lock. These are the expected results, otherwise you can have two
transactions working in the same entity bean at the same time, which completely breaks any
kind of transactional consistency guarantees.
You have two options here:
- Use optimistic locking. This means that each transaction can get its own copy of the
bean.
- Change your code to avoid this dead lock situation.
Note that as the lock scope is tied to the current transaction scope, if you mark the
method that is acquiring the lock as not running in a transaction then the lock will only
be held for the duration of the method invocation.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
9:51 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stephen Coy commented on AS7-5539:
----------------------------------
The closest the old JBoss CMP implementation had to optimistic locking was the
"Instance per Transaction Policy". I wonder if MG has this configured in his
older JBoss versions and forgot about it?
Is this still supported? As I recall, it was quite tricky to build non-trivial CMP
applications without it.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
9:57 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stuart Douglas commented on AS7-5539:
-------------------------------------
It is still supported, if you add <optimisitic-locking> elements to the
jbosscmp-jdbc.xml file.
For an example see
https://github.com/jbossas/jboss-as/tree/master/testsuite/integration/bas...
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12:58 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Hi Stuart,
Thanks for the detailed explanation.
I will try to run the code as suggested by you. But for getting my concept right please
explain the below point:
1. Till JBOSS6 we were using 'Instance Per transaction policy' which is configured
using JBOSS.xml and the policy we have defined in standardjboss.xml. We have not
configured anything in jbosscmp-jdbc.xml. So to achieve the above behavour shall I need to
add this configuration in jbosscmp-jdbc.xml as standardjboss.xml is not available in
JBOSS7?
"For now we are using a completely synchronized approach to entity concurrency. There
is at most 1 entity active for a given primary key at any time, and access is synchronized
within a transaction." this is the comment on EntityBeanSynchronizationInterceptor
2. As you said there is an inner transaction which is trying to acquire the lock on same
entity. But my query is that T1 transaction created an Entity Bean object and it would be
suspended when inner transaction starts as it has been started by RequiresNew, then why it
will be using the entity created in T1 transaction. It should get its own bean. If not,
then what is the default policy configured? And it means that we can never use an object
of same Entity bean in inner transaction if it has been used in a transaction starting
inner transaction?
3. Is the Entity Bean pooling by default supported in JBOSS 7.1.1? There is configuration
available in the XSD of JBOSS EJB, but the default standalone.xml has no such
configuration.
Thanks and Regards,
Mayank
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:04 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/16/12 2:03 PM:
------------------------------------------------------------
Hi Stuart/Stephen,
Thanks for the detailed explanation.
I will try to run the code as suggested by you. But for getting my concept right please
explain the below point:
1. Till JBOSS6 we were using 'Instance Per transaction policy' which is configured
using JBOSS.xml and the policy we have defined in standardjboss.xml. We have not
configured anything in jbosscmp-jdbc.xml. So to achieve the above behavour shall I need to
add this configuration in jbosscmp-jdbc.xml as standardjboss.xml is not available in
JBOSS7?
"For now we are using a completely synchronized approach to entity concurrency. There
is at most 1 entity active for a given primary key at any time, and access is synchronized
within a transaction." this is the comment on EntityBeanSynchronizationInterceptor
2. As you said there is an inner transaction which is trying to acquire the lock on same
entity. But my query is that T1 transaction created an Entity Bean object and it would be
suspended when inner transaction starts as it has been started by RequiresNew, then why it
will be using the entity created in T1 transaction. It should get its own bean. If not,
then what is the default policy configured? And it means that we can never use an object
of same Entity bean in inner transaction if it has been used in a transaction starting
inner transaction?
3. Is the Entity Bean pooling by default supported in JBOSS 7.1.1? There is configuration
available in the XSD of JBOSS EJB, but the default standalone.xml has no such
configuration.
Thanks and Regards,
Mayank
was (Author: ion_mayank):
Hi Stuart,
Thanks for the detailed explanation.
I will try to run the code as suggested by you. But for getting my concept right please
explain the below point:
1. Till JBOSS6 we were using 'Instance Per transaction policy' which is configured
using JBOSS.xml and the policy we have defined in standardjboss.xml. We have not
configured anything in jbosscmp-jdbc.xml. So to achieve the above behavour shall I need to
add this configuration in jbosscmp-jdbc.xml as standardjboss.xml is not available in
JBOSS7?
"For now we are using a completely synchronized approach to entity concurrency. There
is at most 1 entity active for a given primary key at any time, and access is synchronized
within a transaction." this is the comment on EntityBeanSynchronizationInterceptor
2. As you said there is an inner transaction which is trying to acquire the lock on same
entity. But my query is that T1 transaction created an Entity Bean object and it would be
suspended when inner transaction starts as it has been started by RequiresNew, then why it
will be using the entity created in T1 transaction. It should get its own bean. If not,
then what is the default policy configured? And it means that we can never use an object
of same Entity bean in inner transaction if it has been used in a transaction starting
inner transaction?
3. Is the Entity Bean pooling by default supported in JBOSS 7.1.1? There is configuration
available in the XSD of JBOSS EJB, but the default standalone.xml has no such
configuration.
Thanks and Regards,
Mayank
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
5:17 p.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Stuart Douglas commented on AS7-5539:
-------------------------------------
1. You need to add the configuration in jbosscmp-jdbc.xml. Alternatively you can configure
this by default in standalone.xml:
{code}
<subsystem xmlns="urn:jboss:domain:ejb3:1.3">
<entity-bean>
<optimistic-locking enabled="true" />
</entity-bean>
</subsystem>
{code}
2. Because the transaction is suspended, it has not completed. The lock is held until the
transaction either commits or rolls back. If you were using opptimistic locking then you
will get a bean per transaction, and this will not be an issue.
3. see 1.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12:38 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
12:52 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/17/12 1:51 AM:
------------------------------------------------------------
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
was (Author: ion_mayank):
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:02 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/17/12 2:01 AM:
------------------------------------------------------------
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
One thing I cannot understand is why new transaction is getting the entity created in
first transaction. We know there is no concept of child transactions in EJB. Inner
transaction should get its own Entity.
was (Author: ion_mayank):
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:07 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/17/12 2:06 AM:
------------------------------------------------------------
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
One thing I cannot understand is why new transaction is getting the entity created in
first transaction. We know there is no concept of child transactions in EJB. Inner
transaction should get its own Entity. Is it done to ensure that when inner Transaction
completes it might have modified the DB and as a result my entity with T1 transaction
becomes out of sync. and would have stale data?
was (Author: ion_mayank):
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
One thing I cannot understand is why new transaction is getting the entity created in
first transaction. We know there is no concept of child transactions in EJB. Inner
transaction should get its own Entity.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
1:28 a.m.
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta edited comment on AS7-5539 at 9/17/12 2:28 AM:
------------------------------------------------------------
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
One thing I cannot understand is why new transaction is getting the entity created in
first transaction. We know there is no concept of child transactions in EJB. Inner
transaction should get its own Entity. Is it done to ensure that when inner Transaction
completes it might have modified the DB and as a result my entity with T1 transaction
becomes out of sync. and would have stale data?
Klaus Benary also reported the issues with Pessimistic locking in JBOSS7 before
"https://community.jboss.org/message/741026"
was (Author: ion_mayank):
Hi Stuart,
As suggested I put the code optimistic-locking enabled="true" in
standalone.xml.
But still server hangs at the same point.
Next I changed the transaction attribute of all the session beans in the flow as Required
(from RequiresNew), post which
it started working. But still it was hanging which is not always. This intermitent
behavour I can attribute to JMS thread which could still be using RequiresNew and I have
not touched that flow.
This means optimistic-locking enabled="true" setting is not working on the
entities.
Next I can try with enabling it in jbosscmp-jdbc.xml, but I am not able to understand
which configuration to use as you have shown 6-7 different configuration in git example.
But the global configuration on standalone level is not working.
I cannot change the code as my business scenarios could fail because of transaction
policies and this is legacy code where changing transaction attributes like this could
result in dangerous runtime failures.
One thing I cannot understand is why new transaction is getting the entity created in
first transaction. We know there is no concept of child transactions in EJB. Inner
transaction should get its own Entity. Is it done to ensure that when inner Transaction
completes it might have modified the DB and as a result my entity with T1 transaction
becomes out of sync. and would have stale data?
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
midnight
[
https://issues.jboss.org/browse/AS7-5539?page=com.atlassian.jira.plugin.s...
]
Mayank Gupta commented on AS7-5539:
-----------------------------------
Apologies, Optimistic locking is working this way.
But sometimes my transaction works and sometimes it throws "Fetch Out of
Sequence" when EJB access Oracle DB due to which transaction rolls back.
Again this is intermittent.
JBOSS7 server hangs up completely when a EJB2.1 entity bean business
method is called in same transaction in a separate thread from the thread which started
the transaction
----------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Key: AS7-5539
URL: https://issues.jboss.org/browse/AS7-5539
Project: Application Server 7
Issue Type: Bug
Components: EJB, Transactions
Affects Versions: 7.1.1.Final, 7.1.3.Final (EAP), 7.2.0.Alpha1
Reporter: Mayank Gupta
Assignee: jaikiran pai
JBOSS7 server hangs up completely while trying to access a EJB2.1 entity bean method in a
specific scenario.
I am using a session facade pattern in which I call my EntityBean from a Stateless
session bean. For moth the beans I have defined the transaction attribute as REQUIRED.
On debugging the code I found that if the same Entity bean is accessed from two different
threads in same transaction then this issue occurs.
So for the first time my code enters "EntityBeanSynchronizationInterceptor"
code is able to take the lock and assign the requester thread as the lock owner. When
second time it comes to access the same entity bean from different thread in SAME
TRANSACTION (as 2nd required results in joining the transaction), JBOSS7 server code
completely stalls in taking the lock. Moreover this time requester thread is not given the
owner status of the transaction.
Code snippet where code hangs:
final Object lockOwner = getLockOwner(transactionSynchronizationRegistry);
lock.pushOwner(lockOwner);
try {
lock.lock(); //line 80 in EntityBeanSynchronizationInterceptor
and subsequently in:
public void lock() {
if (compareAndSetState(0, 1))
owner = currentRequestor();
else
acquire(1); //Here line 86 in OwnableReentrantLock
}
This is a completely blocking issue for migrating my application from JBOSS6 to JBOSS7.
I get this error in numerous places in my application code.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
4433
days inactive
4439
days old
40 comments
4 participants
participants (4)
-
jaikiran pai (JIRA) -
Mayank Gupta (JIRA)
-
Stephen Coy (JIRA)
-
Stuart Douglas (JIRA)