You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@ignite.apache.org by "Vyacheslav Koptilin (Jira)" <ji...@apache.org> on 2021/11/02 11:45:00 UTC

[jira] [Updated] (IGNITE-15750) Ignite Read Write through with RDBMS based Cache Store

     [ https://issues.apache.org/jira/browse/IGNITE-15750?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]

Vyacheslav Koptilin updated IGNITE-15750:
-----------------------------------------
    Description: 
When we configure ignite as read write through cache in container managed setup, JTA support appears buggy. Expectation and behaviour seen are described below

Given
----
CMT setup with necessary configurations like 

a. JTA transaction manager

b. Tta synchronisation

and 

c. Cache Store based on JDBC drivers

 

Expectations:
 # When JTA transaction commits the write to database should be part of same transaction
 # Similarly behaviour in case of rollbacks

 

Behaviour seen

     1. Ignite write store is initiated post JTA commit 

      This means the write to backend database is not in sync to replicated cache changes  

across cluster.

 

 *UPDATE*: The details and source code are available on the user list: [JTA Ignite Container Managed Transaction Query|https://lists.apache.org/thread/67mwfj9mw4yoljomf2xnh1c5hrjgyp2t]

 

  was:
When we configure ignite as read write through cache in container managed setup, JTA support appears buggy. Expectation and behaviour seen are described below

Given

-----

CMT setup with necessary configurations like 

a. JTA transaction manager

b. Tta synchronisation

and 

c. Cache Store based on JDBC drivers

 

Expectations:
 # When JTA transaction commits the write to database should be part of same transaction
 # Similarly behaviour in case of rollbacks

 

Behaviour seen

     1. Ignite write store is initiated post JTA commit 

      This means the write to backend database is not in sync to replicated cache changes  

across cluster.

 

 

 


> Ignite Read Write through with RDBMS based Cache Store
> ------------------------------------------------------
>
>                 Key: IGNITE-15750
>                 URL: https://issues.apache.org/jira/browse/IGNITE-15750
>             Project: Ignite
>          Issue Type: Bug
>    Affects Versions: 2.10, 2.11
>            Reporter: vikram
>            Priority: Critical
>
> When we configure ignite as read write through cache in container managed setup, JTA support appears buggy. Expectation and behaviour seen are described below
> Given
> ----
> CMT setup with necessary configurations like 
> a. JTA transaction manager
> b. Tta synchronisation
> and 
> c. Cache Store based on JDBC drivers
>  
> Expectations:
>  # When JTA transaction commits the write to database should be part of same transaction
>  # Similarly behaviour in case of rollbacks
>  
> Behaviour seen
>      1. Ignite write store is initiated post JTA commit 
>       This means the write to backend database is not in sync to replicated cache changes  
> across cluster.
>  
>  *UPDATE*: The details and source code are available on the user list: [JTA Ignite Container Managed Transaction Query|https://lists.apache.org/thread/67mwfj9mw4yoljomf2xnh1c5hrjgyp2t]
>  



--
This message was sent by Atlassian Jira
(v8.3.4#803005)