On 8/15/2018 1:26 AM, Derek Poh wrote:
We have a setup of 2 servers, running Solr 6.6.2, on production.
There are 5 collections.
All collection are created as 1 shard x 2 replicas.
4 of the collections have this issue.
A replica of each of this 4 collections is in Recovery Mode. The
affected replicas are on the same server or node.
I noticed there is no Leader node indicated for this 4 collections in
the Solr Admin. This is the screenshot of the Solr Admin
http://imagebucket.net/pmndqkijla5c/solr_admin.PNG This is the
commands I used to stop and start the solr process, bin/solr stop -p
8983 bin/solr start -cloud -p 8983 -s "/apps/search/solr-6.6.2/home"
-z hktszk1:2181,hktszk2:2181,hktszk3:2181 May I know how can I bring
up this replicas? Derek
We see reports of this happening occasionally on different versions.
From what I've seen, you may need to restart ALL of your Solr servers
at least once to clear the problem, even the ones that seem to be
working. It may take more than one restart.
If there continue to be problems, share the solr.log file from each
server so we can look at what the errors are.
Thanks,
Shawn
