[
https://issues.apache.org/jira/browse/IGNITE-16118?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Pavel Pereslegin updated IGNITE-16118:
--------------------------------------
Attachment: IgniteCacheAtomicExpiryPolicyReadTest.java
> Near cache entry can miss a TTL update if an NPE occurs on the primary node.
> ----------------------------------------------------------------------------
>
> Key: IGNITE-16118
> URL: https://issues.apache.org/jira/browse/IGNITE-16118
> Project: Ignite
> Issue Type: Bug
> Reporter: Pavel Pereslegin
> Assignee: Pavel Pereslegin
> Priority: Minor
> Labels: ise
> Fix For: 2.13
>
> Attachments: IgniteCacheAtomicExpiryPolicyReadTest.java,
> ttl-updates.png
>
> Time Spent: 20m
> Remaining Estimate: 0h
>
> In a rare case, you might observe a confusing
> {{GridCacheEntryRemovedException "Failed to send TTL update request"}} in
> logs while *reading a non-expired* cache value.
> {noformat}
> [ERROR][sys-#258%expiry.EntryRemovedOnReadTest2%][root] <default> Failed to
> send TTL update request.
> org.apache.ignite.internal.processors.cache.GridCacheEntryRemovedException
> at
> org.apache.ignite.internal.processors.cache.GridCacheMapEntry.checkObsolete(GridCacheMapEntry.java:3052)
> at
> org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtCacheEntry.checkReadersLocked(GridDhtCacheEntry.java:732)
> at
> org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtCacheEntry.checkReaders(GridDhtCacheEntry.java:708)
> at
> org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtCacheEntry.readers(GridDhtCacheEntry.java:416)
> at
> org.apache.ignite.internal.processors.cache.distributed.dht.GridDhtCacheAdapter$8.run(GridDhtCacheAdapter.java:1122)
> at
> org.apache.ignite.internal.util.IgniteUtils.wrapThreadLoader(IgniteUtils.java:7329)
> at
> org.apache.ignite.internal.processors.closure.GridClosureProcessor$1.body(GridClosureProcessor.java:827)
> at
> org.apache.ignite.internal.util.worker.GridWorker.run(GridWorker.java:125)
> at
> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
> at
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
> at java.lang.Thread.run(Thread.java:748)
> {noformat}
> It looks like there is a race between {{sendTtlUpdateRequest(UUID,
> GridCacheTtlUpdateRequest)}} and heap entry eviction.
> This situation is currently not handled correctly, resulting in a hidden
> NullPointerException that aborts TTL updates.
> Due to the nature of the TTL updates in Ignite, this problem can only occur
> when the TTL is updated from the backup node.
> !ttl-updates.png!
> Thus, in a very rare case, we may notice a lack of updates in the near-cache
> and/or on "non-initiator" backups
--
This message was sent by Atlassian Jira
(v8.20.1#820001)
