[jira] [Commented] (KAFKA-4492) java.lang.IllegalStateException: Attempting to put a clean entry for key... into NamedCache

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/KAFKA-4492?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15729783#comment-15729783
 ] 

ASF GitHub Bot commented on KAFKA-4492:
---

Github user asfgit closed the pull request at:

https://github.com/apache/kafka/pull/2226


> java.lang.IllegalStateException: Attempting to put a clean entry for key... 
> into NamedCache
> ---
>
> Key: KAFKA-4492
> URL: https://issues.apache.org/jira/browse/KAFKA-4492
> Project: Kafka
>  Issue Type: Bug
>  Components: streams
>Affects Versions: 0.10.1.0
>Reporter: Damian Guy
>Assignee: Damian Guy
> Fix For: 0.10.2.0
>
>
> This follows on from https://issues.apache.org/jira/browse/KAFKA-4311
> The exception seems to be triggered by topologies with multiple joined 
> tables. As a new record arrives in one table it triggers an eviction. The 
> eviction causes a flush which will trigger a join processor. These in-turn 
> does a cache lookup and if the value is not in the cache, then it will be 
> retrieved from the store and put in the cache, triggering another eviction. 
> And so on.
> Exception reported on mailing list
> https://gist.github.com/mfenniak/509fb82dfcfda79a21cfc1b07dafa89c
> Further investigation into this also reveals that this same eviction process 
> can send the cache eviction into an infinite loop. It seems that the LRU is 
> broken.



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

[jira] [Commented] (KAFKA-4492) java.lang.IllegalStateException: Attempting to put a clean entry for key... into NamedCache

[ASF GitHub Bot (JIRA)]

[ 
https://issues.apache.org/jira/browse/KAFKA-4492?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel=15729341#comment-15729341
 ] 

ASF GitHub Bot commented on KAFKA-4492:
---

GitHub user dguy opened a pull request:

https://github.com/apache/kafka/pull/2226

KAFKA-4492: java.lang.IllegalStateException: Attempting to put a clean 
entry for key... into NamedCache

The NamedCache wasn't correctly dealing with its re-entrant nature. This 
would result in the LRU becoming corrupted, and the above exception occurring 
during eviction. For example:
Cache A: dirty key 1
eviction runs on Cache A
Node for key 1 gets marked as clean
Entry for key 1 gets flushed downstream
Downstream there is a processor that also refers to the table fronted by 
Cache A
Downstream processor puts key 2 into Cache A
This triggers an eviction of key 1 again ( it is still the oldest node as 
hasn't been removed from the LRU)
As the Node for key 1 is clean flush doesn't run and it is immediately 
removed from the cache. 
So now we have dirtyKey set with key =1, but the value doesn't exist in the 
cache.
Downstream processor tries to put key = 1 into the cache, it fails as key 
=1 is in the dirtyKeySet.

You can merge this pull request into a Git repository by running:

$ git pull https://github.com/dguy/kafka cache-bug

Alternatively you can review and apply these changes as the patch at:

https://github.com/apache/kafka/pull/2226.patch

To close this pull request, make a commit to your master/trunk branch
with (at least) the following in the commit message:

This closes #2226


commit 926ef4b1900c6870819948e12950f215ab3998ac
Author: Damian Guy 
Date:   2016-12-07T14:48:34Z

fix cache bug

commit 6b6f3a57acacae1f0e4642ebc080ec9bc451e98c
Author: Damian Guy 
Date:   2016-12-07T17:12:47Z

add comment about cache eviction




> java.lang.IllegalStateException: Attempting to put a clean entry for key... 
> into NamedCache
> ---
>
> Key: KAFKA-4492
> URL: https://issues.apache.org/jira/browse/KAFKA-4492
> Project: Kafka
>  Issue Type: Bug
>  Components: streams
>Affects Versions: 0.10.1.0
>Reporter: Damian Guy
>Assignee: Damian Guy
> Fix For: 0.10.2.0
>
>
> This follows on from https://issues.apache.org/jira/browse/KAFKA-4311
> The exception seems to be triggered by topologies with multiple joined 
> tables. As a new record arrives in one table it triggers an eviction. The 
> eviction causes a flush which will trigger a join processor. These in-turn 
> does a cache lookup and if the value is not in the cache, then it will be 
> retrieved from the store and put in the cache, triggering another eviction. 
> And so on.
> Exception reported on mailing list
> https://gist.github.com/mfenniak/509fb82dfcfda79a21cfc1b07dafa89c
> Further investigation into this also reveals that this same eviction process 
> can send the cache eviction into an infinite loop. It seems that the LRU is 
> broken.



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