Yes, I've checked them and all nodes are pointing to the sme IP and the
same port (2181). Also all of them are visible in the SolrCloud Graph
section so this would mean they are part of the same cloud.
Largest file is solrconfig which is 58K so this shouldn't be a problem
either.
The potential pro
Hmmm, I didn’t really look carefully at the end of your e-mail. There not being
an /overseer znode _looks_ like one or more of your Solr nodes isn’t connecting
to the proper ZooKeeper ensemble.
bq. All of the instances are able to talk to zookeeper (they are
>
>>> displayed as active in the Sol
Hi Erick, thanks for your reply!
I didn't mention it but we have tried async requests. Then it does not time
out of course, but instead appears to run indefinitely, with REQUESTSTATUS
response like this:
{
"responseHeader":{
"status":0,
"QTime":1},
"status":{
"state":"submitted",
Certainly at times some things just take a long time. The 180
second timeout is fairly arbitrary.
GC pauses, creating a zillion replicas etc. can cause timeouts like
this to be exceeded.
Rather than rely on lengthening some magic timeout value and hoping, I
suggest you use
the async option, se
