Watcher error message - filesystem ran out of space and can't recover


(Dennis) #1


My filesystem ran out of space due to elasticsearch indexes (my fault). Elasticsearch is back up again but I can't do anything with watcher anymore. I get this error in the logs:

[2015-08-06 15:50:58,112][WARN ][index.engine             ] [Contrary] [.triggered_watches][0] failed to sync translog
[2015-08-06 15:50:58,112][WARN ][indices.cluster          ] [Contrary] [[.triggered_watches][0]] marking and sending shard failed due to [failed recovery]
org.elasticsearch.index.gateway.IndexShardGatewayRecoveryException: [.triggered_watches][0] failed to recover shard
        at org.elasticsearch.index.gateway.local.LocalIndexShardGateway.recover(
        at org.elasticsearch.index.gateway.IndexShardGatewayService$
        at java.util.concurrent.ThreadPoolExecutor.runWorker(
        at java.util.concurrent.ThreadPoolExecutor$
Caused by: org.elasticsearch.index.translog.TranslogCorruptedException: translog corruption while reading from stream
        at org.elasticsearch.index.gateway.local.LocalIndexShardGateway.recover(
        ... 4 more
Caused by: org.elasticsearch.ElasticsearchIllegalArgumentException: No version type match [78]
        at org.elasticsearch.index.VersionType.fromValue(
        at org.elasticsearch.index.translog.Translog$Create.readFrom(
        ... 5 more
[2015-08-06 15:50:58,112][WARN ][cluster.action.shard     ] [Contrary] [.triggered_watches][0] received shard failed for [.triggered_watches][0], node[U-KtH2Y1TsCMC9Lp_503lA], [P], s[INITIALIZING], indexUUID [s7VKq9ThTvu5CIhjIrftxw], reason [shard failure [failed recovery][IndexShardGatewayRecoveryException[[.triggered_watches][0] failed to recover shard]; nested: TranslogCorruptedException[translog corruption while reading from stream]; nested: ElasticsearchIllegalArgumentException[No version type match [78]]; ]]

Does anyone have any helpful tips to get it back up and working before i re-install ES and watcher again.

red    open   .triggered_watches                1   1
yellow open   .watches                          1   1          1            2      2.8kb          2.8kb



(Martijn Van Groningen) #2

This is unfortunate. Best way to get Watcher operational again is to delete the .triggered_watches index and wait for Watcher to start automatically or start it via the start api.

(system) #3