[
https://issues.apache.org/jira/browse/OAK-1453?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14074482#comment-14074482
]
Chetan Mehrotra edited comment on OAK-1453 at 7/26/14 6:54 AM:
---------------------------------------------------------------
Couple of observation regarding importance of config servers when sharding is
involved based on this
[thread|https://groups.google.com/d/msg/mongodb-user/Q0yRpr-kNco/DLMtpjZq36IJ]
* you cannot simply randomise the list of config servers per mongos. The
--configdb string needs to be the same across all mongos'.
* Traffic between mongos and config servers can be very high (particularly
first config server) if balancing is going on
* On AWS it might be tempting to use t1.micro for config server but that should
not be done as they might require higher network throughput. Use m3.large
* In prod there should be [*exactly 3* config
servers|http://stackoverflow.com/questions/16232025/why-mongodb-config-servers-must-be-one-or-three-only]
was (Author: chetanm):
Couple of observation regarding importance of config servers when sharding is
involved based on this
[thread|https://groups.google.com/d/msg/mongodb-user/Q0yRpr-kNco/DLMtpjZq36IJ]
* you cannot simply randomise the list of config servers per mongos. The
--configdb string needs to be the same across all mongos'.
* Traffic between mongos and config servers can be very high (particularly
first config server) if balancing is going on
* On AWS it might be tempting to use t1.micro for config server but that should
not be done as they might require higher network throughput. Use m3.large
> MongoMK failover support for replica sets (esp. shards)
> -------------------------------------------------------
>
> Key: OAK-1453
> URL: https://issues.apache.org/jira/browse/OAK-1453
> Project: Jackrabbit Oak
> Issue Type: Improvement
> Components: mongomk
> Reporter: Michael Marth
> Assignee: Thomas Mueller
> Labels: production, resilience
> Fix For: 1.1
>
>
> With OAK-759 we have introduced replica support in MongoMK. I think we still
> need to address the resilience for failover from primary to secoandary:
> Consider a case where Oak writes to the primary. Replication to secondary is
> ongoing. During that period the primary goes down and the secondary becomes
> primary. There could be some "half-replicated" MVCC revisions, which need to
> be either discarded or be ignored after the failover.
> This might not be an issue if there is only one shard, as the commit root is
> written last (and replicated last)
> But with 2 shards the the replication state of these 2 shards could be
> inconsistent. Oak needs to handle such a situation without falling over.
> If we can detect a Mongo failover we could query Mongo which revisions are
> fully replicated to the new primary and discard the potentially
> half-replicated revisions.
--
This message was sent by Atlassian JIRA
(v6.2#6252)
