cluster status is red

[hlyeung]

we have three instances and after migrate the data to another drive due the old drive is almost full, the cluster status is red and there are two Initializing Shards and two unassigned shards. how can i solve it?

[scottwilkerson]

How long have you had two "Initializing Shards"?

Can you run the following and post the results

Code: Select all

curl -XGET http://localhost:9200/_cat/shards | grep INIT

[hlyeung]

it lasted for a few days.


% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
100 82604 100 82604 0 0 106k 0 --:--:-- --:--:-- --:--:-- 106k
logstash-2015.01.28 2 p INITIALIZING 192.168.21.10 181841a1-d717-437c-bd36-6d4a8344abe6
logstash-2015.01.28 0 p INITIALIZING 192.168.20.10 33c658b5-db74-480c-bb71-32c430f77b00

it runned a few days.


IP Hostname Port 1m, 5m, 15m Load CPU % Memory Used Memory Free Storage Total Storage Available Elasticsearch Logstash Actions
192.168.20.10 nls3-tmh2.it.pccwglobal.com 9300 1.32, 1.20, 1.27 34% 62% 37% 249.8GB 92.9GB Elasticsearch is running... Logstash is running... -
192.168.21.10 nls1-hht5.it.pccwglobal.com 9300 0.68, 0.70, 0.70 20% 44% 55% 249.8GB 249.2GB Elasticsearch is running... Logstash is running... -
192.168.22.240 nls1-tmh2.it.pccwglobal.com 9300 0.26, 0.25, 0.30 18% 66% 33% 249.8GB 93.4GB Elasticsearch is running... Logstash is running...

[scottwilkerson]

I would recommend closing the index for logstash-2015.01.28 by going to Administration -> Index Status and click the close icon next to logstash-2015.01.28

If you need data from that index you can verify you have a backup of the index in Administration -> Backup & Maintenance, and if so, you can do the following:
Go to Administration -> Index Status and delete the index logstash-2015.01.28
Go to Administration -> Backup & Maintenance and click restore for index logstash-2015.01.28

Either of these should get you back to green status

[hlyeung]

When i close it, it becomes green.
does that mean we lost the data on that day?
thanks

[scottwilkerson]

When i close it, it becomes green.

This is normal.

does that mean we lost the data on that day?

No, you can restore the index from backup (provided you have one)
or
If you re-open the Index you will be able to query the 3 shards that are active, but the other 2 are likely missing.