[ https://issues.apache.org/jira/browse/IGNITE-4934?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Alexander Belyak resolved IGNITE-4934. -------------------------------------- Resolution: Invalid > GridCacheWriteBehindStore broke if store backend throw a single exception > ------------------------------------------------------------------------- > > Key: IGNITE-4934 > URL: https://issues.apache.org/jira/browse/IGNITE-4934 > Project: Ignite > Issue Type: Bug > Components: cache > Affects Versions: 1.9 > Reporter: Alexander Belyak > Assignee: Alexey Dmitriev > Priority: Critical > Labels: easyfix > Original Estimate: 4h > Remaining Estimate: 4h > > If flusher in GridCacheWriteBehindStore get runtime exception from underlying > CacheStore - it will stop working at all. So future operation with > GridCacheWriteBehindStore will be performed by writer threads (in > flushSingleValue procedure) without batching, without write behind and > moreower - with deadlock if writer will try to owerride some key in pending > state, wich was processed by broken flusher thread. > Reproducer: GridCacheWriteBehindStoreMultithreadedSelfTest.testStoreFailure > with > exp = runPutGetRemoveMultithreaded(10, 10); > changed to > exp = runPutGetRemoveMultithreaded(10, 500); > This test should be changed too. -- This message was sent by Atlassian JIRA (v6.3.15#6346)