Elasticsearch quick recovery after restart


(Ankit Jain) #1

Hi All,

We have 5 nodes cluster and each node serving 100 shards. Elasticsearch
node will take around 10-15 min on restart.

Each shard contains 10 million docs and size of each doc is 2 K.B.

Please suggest some parameters for fast recovery of shards on restart.

Thanks,
Ankit Jain

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Clinton Gormley) #2

Hi Ankit

Read about gateway.recover_after.... settings here:
http://www.elasticsearch.org/guide/reference/modules/gateway/

clint

On 12 September 2013 15:19, Ankit Jain ankitjaincs06@gmail.com wrote:

Hi All,

We have 5 nodes cluster and each node serving 100 shards. Elasticsearch
node will take around 10-15 min on restart.

Each shard contains 10 million docs and size of each doc is 2 K.B.

Please suggest some parameters for fast recovery of shards on restart.

Thanks,
Ankit Jain

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Ankit Jain) #3

Hi clint,

Thanks for reply ...

My problem related to fast recovery (movement from assigned to unassigned
state) all the shards server by dead node during restart.

For example:
Let us consider, node1 contain 100 shards and remaining 4 nodes (node2,
node3, node4 and node5) are in running mode. During recovery (restart) of
node1, it will take around 10-15min to move all shards from unassigned
state to assigned state.

How we can recovery quickly move all the shards from unassigned state to
assigned state?

Regards,
Ankit

On Thursday, 12 September 2013 19:06:58 UTC+5:30, Clinton Gormley wrote:

Hi Ankit

Read about gateway.recover_after.... settings here:
http://www.elasticsearch.org/guide/reference/modules/gateway/

clint

On 12 September 2013 15:19, Ankit Jain <ankitj...@gmail.com <javascript:>>wrote:

Hi All,

We have 5 nodes cluster and each node serving 100 shards. Elasticsearch
node will take around 10-15 min on restart.

Each shard contains 10 million docs and size of each doc is 2 K.B.

Please suggest some parameters for fast recovery of shards on restart.

Thanks,
Ankit Jain

--
You received this message because you are subscribed to the Google Groups
"elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an
email to elasticsearc...@googlegroups.com <javascript:>.
For more options, visit https://groups.google.com/groups/opt_out.

--
You received this message because you are subscribed to the Google Groups "elasticsearch" group.
To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(system) #4