[JBoss JIRA] (ISPN-1862) fire TransactionCompletedEvent not as part of the transaction
by Mircea Markus (JIRA)
[ https://issues.jboss.org/browse/ISPN-1862?page=com.atlassian.jira.plugin.... ]
Mircea Markus updated ISPN-1862:
--------------------------------
Priority: Minor (was: Major)
> fire TransactionCompletedEvent not as part of the transaction
> -------------------------------------------------------------
>
> Key: ISPN-1862
> URL: https://issues.jboss.org/browse/ISPN-1862
> Project: Infinispan
> Issue Type: Enhancement
> Components: Core API
> Affects Versions: 5.1.1.FINAL
> Reporter: Thomas Fromm
> Assignee: Manik Surtani
> Priority: Minor
>
> When catching TransactionCompletedEvent I would expect, the transaction is committed, especially when isTransactionSuccessful() == true.
> I need to perform some operations after a transaction was completed. My workaround is ATM to suspend the transaction, perform my (async) stuff, and
> resume it afterwards. But in case of errors after resume, I get another TransactionCompletedEvent for this transaction with isTransactionSuccessful() == false.
> You see, I need something like a TransactionReallyCompletedEvent :-)
> My change request is, that the TransactionCompletedEvents should be fired, after the transaction was committed.
--
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 years, 9 months
[JBoss JIRA] (ISPN-1862) fire TransactionCompletedEvent not as part of the transaction
by Mircea Markus (JIRA)
[ https://issues.jboss.org/browse/ISPN-1862?page=com.atlassian.jira.plugin.... ]
Mircea Markus updated ISPN-1862:
--------------------------------
Fix Version/s: (was: 6.0.0.Final)
> fire TransactionCompletedEvent not as part of the transaction
> -------------------------------------------------------------
>
> Key: ISPN-1862
> URL: https://issues.jboss.org/browse/ISPN-1862
> Project: Infinispan
> Issue Type: Enhancement
> Components: Core API
> Affects Versions: 5.1.1.FINAL
> Reporter: Thomas Fromm
> Assignee: Manik Surtani
>
> When catching TransactionCompletedEvent I would expect, the transaction is committed, especially when isTransactionSuccessful() == true.
> I need to perform some operations after a transaction was completed. My workaround is ATM to suspend the transaction, perform my (async) stuff, and
> resume it afterwards. But in case of errors after resume, I get another TransactionCompletedEvent for this transaction with isTransactionSuccessful() == false.
> You see, I need something like a TransactionReallyCompletedEvent :-)
> My change request is, that the TransactionCompletedEvents should be fired, after the transaction was committed.
--
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 years, 9 months
[JBoss JIRA] (ISPN-144) graph based deadlock detection
by Mircea Markus (JIRA)
[ https://issues.jboss.org/browse/ISPN-144?page=com.atlassian.jira.plugin.s... ]
Mircea Markus resolved ISPN-144.
--------------------------------
Fix Version/s: (was: 6.0.0.Final)
Resolution: Won't Fix
total order avoids deadlocks entirely.
> graph based deadlock detection
> -------------------------------
>
> Key: ISPN-144
> URL: https://issues.jboss.org/browse/ISPN-144
> Project: Infinispan
> Issue Type: Feature Request
> Reporter: Mircea Markus
> Assignee: Mircea Markus
>
> current deadlock detection impl only knows to look for deadlocks between two transactions: i.e. tx1 (k1,k2) tx (k2,k1). This might be enhanced with transitive deadlock detection check:
> tx1(k1,k2) tx2(k2,k3) tx3(k3,k1). In the previous example, if all transactions are done acquiring first lock we result in a deadlock. Needs further thinking.
--
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 years, 9 months
[JBoss JIRA] (ISPN-2673) maxEntries = 0 support for eviction
by Mircea Markus (JIRA)
[ https://issues.jboss.org/browse/ISPN-2673?page=com.atlassian.jira.plugin.... ]
Mircea Markus commented on ISPN-2673:
-------------------------------------
ISPN-3051 should help with this as well.
> maxEntries = 0 support for eviction
> -----------------------------------
>
> Key: ISPN-2673
> URL: https://issues.jboss.org/browse/ISPN-2673
> Project: Infinispan
> Issue Type: Feature Request
> Components: Eviction
> Affects Versions: 5.2.0.Beta6
> Reporter: Thomas Fromm
> Assignee: Mircea Markus
> Priority: Minor
>
> To save memory I'd like to configure maxEntries to 0.
> In case I use a Cache to share (possible) big data as entry between nodes. These data are persisted inside store and only used rarely. So I see two ways to realize that:
> 1) I can set maxEntries to 0 so that after use the entry is evicted from cache.
> 2) Have a cache mode which does not keep data in memory and just work transparent on the store.
--
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 years, 9 months
[JBoss JIRA] (ISPN-201) make JDBC cache store use DB transactions for commit/rollback
by Mircea Markus (JIRA)
[ https://issues.jboss.org/browse/ISPN-201?page=com.atlassian.jira.plugin.s... ]
Mircea Markus updated ISPN-201:
-------------------------------
Fix Version/s: (was: 6.0.0.Final)
> make JDBC cache store use DB transactions for commit/rollback
> -------------------------------------------------------------
>
> Key: ISPN-201
> URL: https://issues.jboss.org/browse/ISPN-201
> Project: Infinispan
> Issue Type: Feature Request
> Components: Loaders and Stores
> Reporter: Mircea Markus
> Assignee: Tristan Tarrant
> Labels: modshape
>
> current impl of JDBC cache store does not use DB transaction (nor local transactions/Connection.setAutocommit(false)) for storing the modifications created within a transaction. This might leave the store in an inconsistent state one operation fails during commit. This should be changed to either 1. use local transactions or 2. register the driver as an XAResource and rely on the TM to manage the boundaries. I would go for 1, as it would be more consistent with other CStore implementation (different cache store interceptor should be used for managing the 2'nd approach).
--
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 years, 9 months