2:30 p.m.
Dennis Reed created ISPN-2330:
---------------------------------
Summary: JBossMarshaller uses wrong class resolver after stop/start
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app in
EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
2:38 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño updated ISPN-2330:
-----------------------------------
Fix Version/s: 5.2.0.CR1
5.2.0.Final
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:37 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Mircea Markus updated ISPN-2330:
--------------------------------
Fix Version/s: 5.2.0.CR1
(was: 5.2.0.Beta2)
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:14 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Work on ISPN-2330 started by Galder Zamarreño.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:14 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño updated ISPN-2330:
-----------------------------------
Labels: jdg jdg6 (was: )
This should be ported to 5.1.x.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:14 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño commented on ISPN-2330:
----------------------------------------
What's happening here is that the @Inject does not get called again when the cache is
restarted.
The base configuration population should really be happening in an @Start element.
@Inject should only cache the method parameters passed.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:14 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño updated ISPN-2330:
-----------------------------------
Priority: Critical (was: Major)
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
11:42 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow updated ISPN-2330:
-------------------------------
Attachment: AS7clusteringtestServerLogs.zip
Attached are somewhat verbose node0/node1 AS server.log files. I added various INFO
logging/printlns to Infinispan 5.1.x + JBoss Marshalling 1.3.x classes. I don't
expect anyone else to try to read the output but if you want to. node0 is responsible for
marshalling the cache view update with the wrong marshaller and node1 gets confused with
the "java.io.StreamCorruptedException: Unexpected byte found when reading an object:
50"
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
12:26 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
Does the ExtendedRiverMarshaller have a similar problem? That is what we appear to be
using at the time of failure.
The client side uses a ExtendedRiverMarshaller with a ContextClassResolver and the server
side uses a ExtendedRiverMarshaller with a ModularClassResolver.
http://pastie.org/5063200 contains the a few lines from the attached server.log (zip)
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
12:26 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow edited comment on ISPN-2330 at 10/15/12 1:25 PM:
--------------------------------------------------------------
Does the ExtendedRiverMarshaller have a similar problem? That is what we appear to be
using at the time of failure ("java.io.StreamCorruptedException: Unexpected byte
found when reading an object: 50"). It is possible that I'm seeing a different
issue.
The client side uses a ExtendedRiverMarshaller with a ContextClassResolver and the server
side uses a ExtendedRiverMarshaller with a ModularClassResolver.
http://pastie.org/5063200 contains the a few lines from the attached server.log (zip)
was (Author: smarlow):
Does the ExtendedRiverMarshaller have a similar problem? That is what we appear to be
using at the time of failure.
The client side uses a ExtendedRiverMarshaller with a ContextClassResolver and the server
side uses a ExtendedRiverMarshaller with a ModularClassResolver.
http://pastie.org/5063200 contains the a few lines from the attached server.log (zip)
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
1:22 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Dennis Reed commented on ISPN-2330:
-----------------------------------
Scott, your issue does appear to be the same as this one. The error messages are
identical to some I've seen from this issue.
Note that it's not ExtendedRiverMarshaller that has the issue, it's the wrong
class resolver being passed in to JBoss Marshalling
higher up in the stack (org.infinispan.marshall.jboss.AbstractJBossMarshaller).
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
1:24 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Dennis Reed commented on ISPN-2330:
-----------------------------------
Further digging shows the issue appears to be that the wrong JBossMarshaller is used,
after it's been stopped.
A new instance is created and inject is called on it, but the instance from the first
start is still used.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
4:19 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
For my issue, stop/start methods don't appear to be involved.
http://pastebin.com/QiGV7F3b shows us creating a new
ExtendedRiverMarshaller/ContextClassResolver pair which is used right away for
marshalling.
I do see some calls to JBossMarshaller.stop() during my testing but that is very early in
the test run (well before any of this occurs and well after the failures occur).
I'll copy this same to the infinispan-dev ml and stop commenting on this thread unless
it indeed does turn out to be the same issue somehow.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
6:15 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow updated ISPN-2330:
-------------------------------
Attachment: anotherAS7clusteringtestServerLogs.zip
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
6:49 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
I attached new logs (anotherAS7clusteringtestServerLogs.zip) to help determine if I'm
also hitting a form of ISPN-2330.
Search for "BANG org.infinispan.util.ImmutableListCopy case hit for" in
jbossas-clustering-SYNC-tcp-0/standalone/log/server.log, as that is when we are actually
going to marshall with the wrong class resolver (a ContextClassResolver). The first one
reached is:
{quote}
16:55:38,187 INFO [stdout] (transport-thread-15) BANG
org.infinispan.util.ImmutableListCopy case hit for AbstractClassResolver
this=org.jboss.marshalling.ContextClassResolver@2a73f761
{quote}
Search backwards for the constructor of ContextClassResolver@2a73f761 and note the call
stack.
{quote}
16:55:27,182 INFO [stdout] (transport-thread-15) ContextClassResolver ctor for
org.jboss.marshalling.ContextClassResolver@2a73f761
{quote}
Look ahead at the marshaller ctor, you will soon see it (same call stack as above
originating from
org.infinispan.marshall.jboss.AbstractJBossMarshaller$PerThreadInstanceHolder.getMarshaller(AbstractJBossMarshaller.java:348)):
{quote}
16:55:27,186 INFO [stdout] (transport-thread-15) xxx AbstractMarshaller ctor is using
classresolver of org.jboss.marshalling.ContextClassResolver ,
this=org.infinispan.marshall.jboss.ExtendedRiverMarshaller@3cecc1e1
{quote}
You will soon see the log message from the AbstractMarshaller ctor that shows the
marshaller + class resolver together:
{quote}
16:55:27,188 INFO [stdout] (transport-thread-15) xxx AbstractMarshaller ctor using
lassResolver org.jboss.marshalling.ContextClassResolver@2a73f761 ,
this=org.infinispan.marshall.jboss.ExtendedRiverMarshaller@3cecc1e1
{quote}
Repeating the first "BANG" from above, note that if find the thead dump below
this line that is for thread transport-thread-15, it is triggered from
org.jboss.marshalling.AbstractClassResolver.annotateClass and shows the same
ContextClassResolver@2a73f761 is being used.
{quote}
16:55:38,187 INFO [stdout] (transport-thread-15) BANG
org.infinispan.util.ImmutableListCopy case hit for AbstractClassResolver
this=org.jboss.marshalling.ContextClassResolver@2a73f761
{quote}
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
https://github.com/scottmarlow/infinispan/tree/5.1.x_hexdump contains my logging changes
to show more information (in case that is useful to others).
I'll also attach a new moreAS7clusteringtestServerLogs.zip which includes logging
changes in JBossMarshaller (start/stop/inject methods).
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
9:53 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow updated ISPN-2330:
-------------------------------
Attachment: moreAS7clusteringtestServerLogs.zip
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
4:56 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño commented on ISPN-2330:
----------------------------------------
The root cause appears to be the use of the old cache marshaller instance after cache
restart. Indirectly, Dan has fixed this issue in ISPN-2297 (2nd issue), and I've added
a patch on top of it so that it works as expected for cache restarts. So, can you please
give it a shot to the latest of my branch in
https://github.com/galderz/infinispan/tree/t_2330_5 and see if the issue is gone?
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:10 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Radoslav Husar commented on ISPN-2330:
--------------------------------------
Thanks Galder, I am going to try this now.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
6:50 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Radoslav Husar updated ISPN-2330:
---------------------------------
Fix Version/s: 5.1.8.Final
Galder your branch *does* fix the issue, which is great news! Thanks.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
7:48 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
^ sounds very encouraging! I merged the fixes in and am building locally to test on my
machine as well. I'll post here when I have results.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
8:17 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Scott Marlow commented on ISPN-2330:
------------------------------------
org.jboss.as.test.clustering.cluster.ejb3.stateful.StatefulFailoverTestCase passed on my
machine, which previously failed 100% of the time. I'm using
https://github.com/galderz/infinispan/tree/t_2330_5 and a zillion local logging changes.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
9:54 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Galder Zamarreño updated ISPN-2330:
-----------------------------------
Status: Pull Request Sent (was: Coding In Progress)
Git Pull Request: https://github.com/infinispan/infinispan/pull/1400,
https://github.com/infinispan/infinispan/pull/1401
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
5:43 p.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Dennis Reed commented on ISPN-2330:
-----------------------------------
I can also confirm that https://github.com/galderz/infinispan/tree/t_2330_5 fixes all my
issues with redeploying clustered apps in EAP6.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
7:38 a.m.
[
https://issues.jboss.org/browse/ISPN-2330?page=com.atlassian.jira.plugin....
]
Dan Berindei updated ISPN-2330:
-------------------------------
Status: Resolved (was: Pull Request Sent)
Resolution: Done
Integrated on master and on 5.1.x.
JBossMarshaller uses wrong class resolver after stop/start
----------------------------------------------------------
Key: ISPN-2330
URL: https://issues.jboss.org/browse/ISPN-2330
Project: Infinispan
Issue Type: Bug
Components: Marshalling
Affects Versions: 5.1.4.FINAL
Reporter: Dennis Reed
Assignee: Galder Zamarreño
Priority: Critical
Labels: jdg, jdg6
Fix For: 5.2.0.CR1, 5.2.0.Final, 5.1.8.Final
Attachments: anotherAS7clusteringtestServerLogs.zip,
AS7clusteringtestServerLogs.zip, moreAS7clusteringtestServerLogs.zip
org.infinispan.marshall.jboss.JBossMarshaller initializes the classResolver in its
inject() method and clears it in its stop() method.
If the cache is stopped and restarted (for example when redeploying a clustered web app
in EAP), the wrong class resolver is used.
Either the classResolver should not be removed in stop() (testing with it removed did not
show any class leaking issues), or it should be reset in start().
--
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
4234
days inactive
4261
days old
infinispan-issues@lists.jboss.org
24 comments
6 participants
participants (6)
-
Dan Berindei (JIRA) -
Dennis Reed (JIRA)
-
Galder Zamarreño (JIRA)
-
Mircea Markus (JIRA)
-
Radoslav Husar (JIRA)
-
Scott Marlow (JIRA)