Elasticsearch recovering process took a long time


(silvina) #1

Hi,
We are using elasticsearch 0.90.1. We had some problems with the network,
when the network is down (the convergence time is around 40s). Recovering
the elastic after this event took a long time to be available. We have 16
data nodes and 16 shards with 2 replica and the settings:

discovery.zen.minimum_master_nodes: 4
gateway.recover_after_nodes: 4
gateway.expected_nodes: 6

Any idea about how to minimize the recovery time?
Is it a good idea to update the version?
What will happen if we increase discovery.zen.fd.ping_interval and
ping_timeout settings? Assuming the network is completely off, is there any
way to wait for at least those 40 seconds to start marking servers down?
Does the ping_timeout ignore the failure to connect to a node?

Thanks.

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/a32a0c42-122a-45a3-9c56-8869b4028608%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


(silvina) #2

Sorry, the setting discovery.zen.minimum_master_nodes is 8

El lunes, 21 de julio de 2014 17:20:03 UTC-3, anivlis escribió:

Hi,
We are using elasticsearch 0.90.1. We had some problems with the network,
when the network is down (the convergence time is around 40s). Recovering
the elastic after this event took a long time to be available. We have 16
data nodes and 16 shards with 2 replica and the settings:

discovery.zen.minimum_master_nodes: 4
gateway.recover_after_nodes: 4
gateway.expected_nodes: 6

Any idea about how to minimize the recovery time?
Is it a good idea to update the version?
What will happen if we increase discovery.zen.fd.ping_interval and
ping_timeout settings? Assuming the network is completely off, is there any
way to wait for at least those 40 seconds to start marking servers down?
Does the ping_timeout ignore the failure to connect to a node?

Thanks.

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/7997de22-b421-4f7c-ae7f-e001dbfef618%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.


(Ivan Brusic) #3

Recovery is throttled since version 0.90.1

http://www.elasticsearch.org/guide/en/elasticsearch/reference/0.90/index-modules-store.html#store-throttling

Increase indices.store.throttle.max_bytes_per_sec to a level that is
suitable for your environment. Since IO should be the main bottleneck, the
setting could vary greatly depending on SSD, platter disk or shared storage.

--
Ivan

On Mon, Jul 21, 2014 at 1:28 PM, anivlis svlucero@gmail.com wrote:

Sorry, the setting discovery.zen.minimum_master_nodes is 8

El lunes, 21 de julio de 2014 17:20:03 UTC-3, anivlis escribió:

Hi,
We are using elasticsearch 0.90.1. We had some problems with the network,
when the network is down (the convergence time is around 40s). Recovering
the elastic after this event took a long time to be available. We have 16
data nodes and 16 shards with 2 replica and the settings:

discovery.zen.minimum_master_nodes: 4
gateway.recover_after_nodes: 4
gateway.expected_nodes: 6

Any idea about how to minimize the recovery time?
Is it a good idea to update the version?
What will happen if we increase discovery.zen.fd.ping_interval and
ping_timeout settings? Assuming the network is completely off, is there any
way to wait for at least those 40 seconds to start marking servers down?
Does the ping_timeout ignore the failure to connect to a node?

Thanks.

--
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.
To view this discussion on the web visit
https://groups.google.com/d/msgid/elasticsearch/7997de22-b421-4f7c-ae7f-e001dbfef618%40googlegroups.com
https://groups.google.com/d/msgid/elasticsearch/7997de22-b421-4f7c-ae7f-e001dbfef618%40googlegroups.com?utm_medium=email&utm_source=footer
.

For more options, visit https://groups.google.com/d/optout.

--
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.
To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/CALY%3DcQACJs%3Dzuu4gjYTvcKpF%2BSgv_1UkYZw%3DC0G8FkY1meAg-Q%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.


(system) #4