[jira] [Commented] (IGNITE-1151) Cache operations can throw ClusterTopologyException without wrapping it into CacheException

["Dmitriy Setrakyan (JIRA)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/IGNITE-1151?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14639303#comment-14639303 ] 

Dmitriy Setrakyan commented on IGNITE-1151:
-------------------------------------------

I think we can create a simple test for it:

- thread 1 constantly doing cache operations
- thread 2 constantly starting and stopping nodes

> Cache operations can throw ClusterTopologyException without wrapping it into CacheException
> -------------------------------------------------------------------------------------------
>
>                 Key: IGNITE-1151
>                 URL: https://issues.apache.org/jira/browse/IGNITE-1151
>             Project: Ignite
>          Issue Type: Bug
>          Components: cache
>    Affects Versions: 1.1.4
>            Reporter: Valentin Kulichenko
>              Labels: Usability
>
> This was reported by one of our users who was testing cache transactions and other operation with changing topology. He claims that he had to catch {{ClusterTopologyException}} directly in some cases, so he had to write the code like this:
> {code}
> try (final Transaction tx = txn.txStart()) {
>     ...
> }
> catch (final CacheException e) {
>     if (!(e.getCause() instanceof ClusterTopologyException)) {
>         throw e;
>     }
>     // retry
>     failure = e;
> }
> catch (final ClusterTopologyException e) {
>     // retry
>     failure = e;
> }
> {code}
> Cache operations should throw only {{CacheException}}. Need to revisit exception handling logic.



--
This message was sent by Atlassian JIRA
(v6.3.4#6332)