Will do!
Thanks!
On Fri, Oct 8, 2010 at 4:16 PM, Shay Banon shay.banon@elasticsearch.comwrote:
First, it does not sound like exactly the same problem. And the 25k number
is very bogus, don't know where you came up with it. Its like saying, it
seems like once I go out with a hat, it seems to rain, hence, me wearing a
hat means its going to rain.Please open another thread on the problems you have. Most times, a node
will break when it runs out of memory due to capacity problems. When you
open that thread, point at any exceptions you get when the node fails.-shay.banon
On Fri, Oct 8, 2010 at 8:59 PM, Pablo Borges pablort@gmail.com wrote:
I have exactly the same problem with Local Storage and 0.11 version os ES.
The only way to recover is to reindex the data.
The problem is, when the index is on that state, you cannot DELETE it, so
I have to reindex all indexes.I also found out this problem occurs frequently on big indexes (big as in
number of documents). Anything higher than 25k items is subject to this
problem.I've removed the sharding (shard: 1) of all indexes, but the problem
remains, so I have to throttle the writing and hope for the best, but I'm
still seeing a crash everyday (sometimes twice a day).On Fri, Oct 8, 2010 at 6:27 AM, Shay Banon shay.banon@elasticsearch.comwrote:
You can try and reindex the data on the 0.9 one.
On Fri, Oct 8, 2010 at 2:37 AM, John Chang jchangkihtest2@gmail.comwrote:
I think we have hit the S3-related recovery bugs in 0.9 you mentioned
above.--All the nodes that were in INITIALIZING state hours ago are still
INITIALIZING.--The disk usage on the work dir of the cluster machines is not growing
(it's been the same for hours), so I don't think anything more is being
downloaded. Our total index volume is only 3.5 gigs.--One of the cluster machines has been logging this over and over for
hours:
[00:24:05,885][DEBUG][index.shard.service ] [Scarlet Beetle]
[index0][3] state: [CREATED]->[RECOVERING]
[00:24:05,886][DEBUG][index.shard.service ] [Scarlet Beetle]
[index0][3] state: [RECOVERING]->[CREATED], restored after recoverySo, I am actively working on upgrading to 0.11, but in the mean time, is
there a work-around to get the cluster up and responsive? Thanks.View this message in context:
http://elasticsearch-users.115913.n3.nabble.com/BroadcastShardOperationFailedException-tp1650661p1651779.html
Sent from the Elasticsearch Users mailing list archive at Nabble.com.
