Home | About | Sematext search-lucene.com search-hadoop.com
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB
 Search Hadoop and all its subprojects:

Switch to Threaded View
HBase >> mail # dev >> replicating to oneself?


Copy link to this message
-
Re: replicating to oneself?
Are you re-deploying over an existing installation? Is it your intention to
preserve data between deployments or are you running in a testing
environment? Are you clearing ZK as part of deploying a fresh cluster or
are you re-using existing znodes? How did you configure replication in the
shell? Can you provide those commands? I'd request debug logs from
o.a.h.h.regionserver.Replication but i don't see much logging in there
anyway.

Basically, can you repro this in a fresh deployment? As Himanshu points
out, I'm suspect of stale configuration hanging around.
On Thu, Oct 31, 2013 at 8:02 PM, Demai Ni <[EMAIL PROTECTED]> wrote:

> Nick,
>
> thanks for looking into this problem. I attached the hbase-site.xml in
> this email. Just like to point out that I have to tear down the cluster I
> posted the original log. so the hbase-site.xml is from another
> cluster(single-node) with the same problem.
>
> BTW, I did some investigation this afternoon and don't  think this is a
> problem within hbase code. (background: I am working within a software
> team, and quite a few engineers change hbase, hadoop, and other codes
> everyday)I tried out several different installations, and found out a week
> ago's build with today's hbase build work just fine; but today's build with
> last week's hbase doesn't. Our build includes hadoop 2, which can introduce
> something problematic.
>
> wondering how hbase generate UUID? maybe that is something I should look
> into? thanks
>
> Demai
>
>
>
>
>
> On Thu, Oct 31, 2013 at 6:20 PM, Nick Dimiduk <[EMAIL PROTECTED]> wrote:
>
>> Can you post your replication settings from hbase-site.xml?
>>
>> On Thursday, October 31, 2013, Demai Ni wrote:
>>
>> > hi, folks,
>> >
>> > I got a strange thing happening on my cluster(hbase 0.94.9) recently. I
>> am
>> > setting up a new cluster for replication, and didn't see the data being
>> > replicated over the peer. Then, I found the following in the log of the
>> > regionserver of the Master:
>> >
>> > 2013-10-31 13:33:03,293 INFO org.apache.hadoop.hbase.metrics: new
>> MBeanInfo
>> > 2013-10-31 13:33:03,300 INFO
>> > org.apache.hadoop.hbase.replication.regionserver.ReplicationSource:
>> Getting
>> > 1 rs from peer cluster # 3
>> > 2013-10-31 13:33:03,300 INFO
>> > org.apache.hadoop.hbase.replication.regionserver.ReplicationSource:
>> > Choosing peer hdtest018.svl.ibm.com,60020,1383251582072
>> > 2013-10-31 13:33:03,302 INFO
>> > org.apache.hadoop.hbase.replication.regionserver.ReplicationSource:
>> > Replicating *b520de1d-3a18-4aec-bd45-de000e81417d* -> *
>> > b520de1d-3a18-4aec-bd45-de000e81417d*
>> >
>> > the log is from ReplicationSource:
>> > *LOG.info("Replicating "+clusterId + " -> " + peerClusterId);*
>> >
>> > It seems the problematic cluster is replicating to itself.
>> > Any suggestion about how to look into this problem? Many thanks
>> >
>> > BTW, I can replicate from another cluster to this problematic one.
>> >
>> > Demai
>> >
>>
>
>
NEW: Monitor These Apps!
elasticsearch, apache solr, apache hbase, hadoop, redis, casssandra, amazon cloudwatch, mysql, memcached, apache kafka, apache zookeeper, apache storm, ubuntu, centOS, red hat, debian, puppet labs, java, senseiDB