ES data on Glusterfs


(Josh Harrison) #1

Is it a dumb idea to have my ES data directory on a glusterFS/lusterfS
storage node? Because we've got some BIG data that we'd like to index, but
it's too big for the local storage on our test cluster.

--
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/11c1be2e-a908-454e-92b5-e65f90e33156%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Jörg Prante) #2

Use Gluster on its native protocol, not on NFS and the like.

If you want backup/restore, Gluster support was announced as a post 1.0 ES
feature.

Jörg

--
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/CAKdsXoGcB0OvxWr48L%2BwSJ3EgBdXnxbYaaSezA%3Dzw-ahQbYvdw%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Josh Harrison) #3

Cool, thanks. It looks like I've conflated gluster and lustre which are
unfortunately totally unrelated. We're running lustre.

On Tuesday, December 17, 2013 5:58:39 PM UTC-8, Jörg Prante wrote:

Use Gluster on its native protocol, not on NFS and the like.

If you want backup/restore, Gluster support was announced as a post 1.0 ES
feature.

Jörg

--
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/212a220f-33eb-4026-9454-1d4b84e2ad17%40googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Jörg Prante) #4

Same with Lustre. Do not use it over the network, because then it is slow.

Also, with ES replica shards and local gateway with recovery over the
network, you have a mechanism similar to what you can achieve with Lustre,
so there is not much sense in duplicating it. YMMV.

Just my 2p.

Jörg

--
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/CAKdsXoH8e8N0y6j97STL2NWFd-QC%2B4wx7Q3ea-%2BRGs-vYuK-FA%40mail.gmail.com.
For more options, visit https://groups.google.com/groups/opt_out.


(Abhijit Paul) #5

@Josh_Harrison @jprante Hello folks, I am trying out to setup for "ES data on Glusterfs" but facing issues like, "CorruptIndexException", and due to that ES index health trued RED, cache consistency issue and Openshift Logging ElasticSearch FSLocks when using GlusterFS storage backend
All this issue are facing with glusterfs 3.10.1 version.
Any pointers or forwards will be appreciated.


(system) #6