[
https://issues.apache.org/jira/browse/ZOOKEEPER-2145?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14715496#comment-14715496
]
Camille Fournier commented on ZOOKEEPER-2145:
---------------------------------------------
Would love to look into this but it will be hard to reproduce without
transaction logs and snapshots. Any chance anyone has an instance with more
detailed information they can share?
> Node can be seen but not deleted
> --------------------------------
>
> Key: ZOOKEEPER-2145
> URL: https://issues.apache.org/jira/browse/ZOOKEEPER-2145
> Project: ZooKeeper
> Issue Type: Bug
> Affects Versions: 3.4.6
> Reporter: Frans Lawaetz
>
> I have a three-server ensemble that appears to be working fine in every
> respect but for the fact that I can ls or get a znode but can not rmr it.
> >[zk: localhost:2181(CONNECTED) 0] get
> >/accumulo/9354e975-7e2a-4207-8c7b-5d36c0e7765d/masters/goal_state
> CLEAN_STOP
> cZxid = 0x15
> ctime = Fri Feb 20 13:37:59 CST 2015
> mZxid = 0x72
> mtime = Fri Feb 20 13:38:05 CST 2015
> pZxid = 0x15
> cversion = 0
> dataVersion = 2
> aclVersion = 0
> ephemeralOwner = 0x0
> dataLength = 10
> numChildren = 0
> [zk: localhost:2181(CONNECTED) 1] rmr
> /accumulo/9354e975-7e2a-4207-8c7b-5d36c0e7765d/masters/goal_state
> Node does not exist:
> /accumulo/9354e975-7e2a-4207-8c7b-5d36c0e7765d/masters/goal_state
> I have run a 'stat' against all three servers and they seem properly
> structured with a leader and two followers. An md5sum of all zoo.cfg shows
> them to be identical.
> The problem seems localized to the accumulo/935.... directory as I can create
> and delete znodes outside of that path fine but not inside of it.
> For example:
> [zk: localhost:2181(CONNECTED) 12] create
> /accumulo/9354e975-7e2a-4207-8c7b-5d36c0e7765d/fubar asdf
> Node does not exist: /accumulo/9354e975-7e2a-4207-8c7b-5d36c0e7765d/fubar
> [zk: localhost:2181(CONNECTED) 13] create /accumulo/fubar asdf
> Created /accumulo/fubar
> [zk: localhost:2181(CONNECTED) 14] ls /accumulo/fubar
> []
> [zk: localhost:2181(CONNECTED) 15] rmr /accumulo/fubar
> [zk: localhost:2181(CONNECTED) 16]
> Here is my zoo.cfg:
> tickTime=2000
> initLimit=10
> syncLimit=15
> dataDir=/data/extera/zkeeper/data
> clientPort=2181
> maxClientCnxns=300
> autopurge.snapRetainCount=10
> autopurge.purgeInterval=1
> server.1=cdf61:2888:3888
> server.2=cdf62:2888:3888
> server.3=cdf63:2888:3888
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
