Hello,
Yesterday I had to restart my server. After loading I checked that elastic-search is running and found that 4 of are "INITIALIZING" stat.
My configuration is built with:
- Elastic search version 2.3
- 5 shards hosted on same server ("node-01").
- 8 CPU\30GB RAM
- less then 10GB of data.
Result of GET "/_cat/shards":
index shard prirep state ip node
MyIndex 4 p INITIALIZING 127.0.0.1 node-01
MyIndex 3 p INITIALIZING 127.0.0.1 node-01
MyIndex 1 p INITIALIZING 127.0.0.1 node-01
MyIndex 2 p UNASSIGNED
MyIndex 0 p INITIALIZING 127.0.0.1 node-01
.kibana 0 p UNASSIGNED
As you can see, "MyIndex->Shard #2" is unassigned - so I called POST "/_cluster/reroute" (from "node-01" to "node-01").
From my previous experience, this initialize loading is takes a lot of time. So I leave the machine working and checked again after 12 hours.
After 12 hours the results of "/_cat/shards" was the same (aka 4 initializing, 1 unassigned). At this point, I was expecting to get better results.
Here it come even more strange. I called GET "/_cat/recovery" and ES respond 200 but the list is empty. That's mean that nothing is happen?
What do you recommend me to do? Why the recovery list is empty if the shards are "INITIALIZING"?