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
Zookeeper >> mail # user >> Rolling config change considered harmful?


Copy link to this message
-
RE: Rolling config change considered harmful?
Not sure if this helps but here is an example:

- Txn T is acknowledged by A and B (ensemble is {A, B, C})
- Ensemble changes to {B, C, D}
- C and D form a quorum and elect C because it has the highest zxid.

C won't have T, so the txn gets lost.

Does it make sense?

-Flavio

> -----Original Message-----
> From: Jordan Zimmerman [mailto:[EMAIL PROTECTED]]
> Sent: 14 June 2013 19:35
> To: [EMAIL PROTECTED]
> Subject: Rolling config change considered harmful?
>
> I'm trying to track a race issue in our ZK usage. In debugging, I've come
to
> wonder if the way we do rolling config ensemble changes is the culprit.
How
> does ZK internal leader election work if the list of servers is not
consistent on
> each instance. e.g. Given 3 instances A B C and new instance coming online
D.
> If A and B think the server list is  (A, B, C) but C and D think the
server list is (B,
> C, D) how does ZooKeeper handle leader election?
>
> -Jordan
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