[
https://issues.apache.org/jira/browse/IGNITE-1233?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14696786#comment-14696786
]
Andrey Gura commented on IGNITE-1233:
-------------------------------------
I seems that preloader demand worker create entries with old values. See debug
info bellow.
For all problem entries there is the same sequence of events when after remove
from some system pool thread occures invoke of {{entry0}} method with {{create
== true}} flag from preloader demand worker.
{noformat}
ts=1439545326988 method=rmv
nodeId=000ccddc-6ccd-45c7-8225-c72053a10000 thread=async-runner-1
ts=1439545326988 method=GCA:entryEx(k top)
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#196%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=GCA:entry0:create=true
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#196%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=GCA:entryEx(k touch)
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#196%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=GCA:entry0:create=true
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#196%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=GCA:entryEx(k top)
nodeId=208ff144-29d3-46e2-8a69-f069db0b4002
thread=ignite-#122%sys-dht.GridCacheDhtClientRemoveFailureTest2%
ts=1439545326988 method=GCA:entry0:create=true
nodeId=208ff144-29d3-46e2-8a69-f069db0b4002
thread=ignite-#122%sys-dht.GridCacheDhtClientRemoveFailureTest2%
ts=1439545326988 method=GCA:entryEx(k touch)
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#197%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=GCA:entry0:create=true
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#197%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=iRmv
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#198%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=removeEntry
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=ignite-#198%sys-dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545326988 method=iRmv
nodeId=208ff144-29d3-46e2-8a69-f069db0b4002
thread=ignite-#123%sys-dht.GridCacheDhtClientRemoveFailureTest2%
ts=1439545326988 method=removeEntry
nodeId=208ff144-29d3-46e2-8a69-f069db0b4002
thread=ignite-#123%sys-dht.GridCacheDhtClientRemoveFailureTest2%
ts=1439545327029 method=GCA:entryEx(k touch)
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=preloader-demand-worker-#235%dht.GridCacheDhtClientRemoveFailureTest1%
ts=1439545327029 method=GCA:entry0:create=true
nodeId=10103a1e-cf93-4ed4-aa11-6db9daea7001
thread=preloader-demand-worker-#235%dht.GridCacheDhtClientRemoveFailureTest1%
{noformat}
> Some tests derived from GridCacheAbstractRemoveFailureTest class fail
> sometimes
> -------------------------------------------------------------------------------
>
> Key: IGNITE-1233
> URL: https://issues.apache.org/jira/browse/IGNITE-1233
> Project: Ignite
> Issue Type: Bug
> Components: cache
> Reporter: Andrey Gura
> Assignee: Andrey Gura
>
> Some tests derived from {{GridCacheAbstractRemoveFailureTest}} class fail
> sometimes. Most stable failures thre are in the following tests (accordingly
> to failure probability):
> *
> {{org.apache.ignite.internal.processors.cache.distributed.dht.GridCacheDhtClientRemoveFailureTest}}
> *
> {{org.apache.ignite.internal.processors.cache.distributed.near.GridCacheNearRemoveFailureTest}}
> *
> {{org.apache.ignite.internal.processors.cache.distributed.dht.GridCacheDhtRemoveFailureTest}}
> This problem was found during fixing IGNITE-1175 ticket. It seams that there
> is concurrency issue that leads to failures.
> Log snippet:
> {noformat}
> [20:14:22,624][ERROR][test-runner][root] Unexpected cache data
> [exp=7923=GridTuple [val=null], actual=55, nodePrimary=false,
> nodeBackup=false, nodeIdx0, nodeId=001ba37c-455a-4254-81b9-bd4e20cb2000]
> [20:14:23,078][ERROR][test-runner][root] Unexpected cache data
> [exp=7923=GridTuple [val=null], actual=55, nodePrimary=true,
> nodeBackup=false, nodeIdx1, nodeId=1089c623-6330-4b6a-9bd7-b5217b63a001]
> [20:14:23,322][ERROR][test-runner][root] Unexpected cache data
> [exp=7923=GridTuple [val=null], actual=55, nodePrimary=false,
> nodeBackup=true, nodeIdx2, nodeId=20690014-336b-4dc0-8e2d-3f173e68e002]
> {noformat}
> Of course the nodes list depends on current topology and can varying from 1
> to 3 for this tests.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
