[ 
https://issues.apache.org/jira/browse/ZOOKEEPER-2248?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14698164#comment-14698164
 ] 

Yongcheng Liu commented on ZOOKEEPER-2248:
------------------------------------------

logCount in SyncRequestProcessor,  change from local variable to global 
variable can solve this problem .

> log.zxid file is very large size. It's likely to lead to zk disk space is not 
> enough.
> -------------------------------------------------------------------------------------
>
>                 Key: ZOOKEEPER-2248
>                 URL: https://issues.apache.org/jira/browse/ZOOKEEPER-2248
>             Project: ZooKeeper
>          Issue Type: Improvement
>            Reporter: Yongcheng Liu
>
> 1. about logCount(the count of log entries)  
> 2. it's local variable in SyncRequestProcessor, it will be reset 0, when 
> SyncRequestProcessor thread exit for LOOKING.   SyncRequestProcessor thread 
> will exit but zk process is ok, when follower followWithLeader and connecting 
> is broken.
> 3. it will lead to log.zxid file keep expanding 64M



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

Reply via email to