Oh great, thanks for the hint!
I've upvoted this issue, since I think it might be worth to be able to
configure that (rather low) ThreadPool count.
On Wed, 3 Apr 2019 at 10:23, Shalin Shekhar Mangar
wrote:
> Thanks Roger. This was reported earlier but missed our attention.
>
> The issue is
Thanks Roger. This was reported earlier but missed our attention.
The issue is https://issues.apache.org/jira/browse/SOLR-11208
On Tue, Apr 2, 2019 at 5:56 PM Roger Lehmann
wrote:
> To be more specific: I currently have 19 collections, where each node has
> exactly one replica per collection.
To be more specific: I currently have 19 collections, where each node has
exactly one replica per collection. A new node will automatically create
new replicas on itself, one for each existing collection (see
cluster-policy above).
So when removing a node, all 19 collection replicas of it need to
Situation
I'm currently trying to set up SolrCloud in an AWS Autoscaling Group, so
that it can scale dynamically.
I've also added the following triggers to Solr, so that each node will have
1 (and only one) replication of each collection:
{
"set-cluster-policy": [
{"replica": "<2", "shard":
