Hi, We have the following setup of Solr-8.11.0 running on an AWS EC2 instance.
- Number of shards: 48 [Without replicas] - Number of external zookeeper : 1 - Index is on an attached EBS on the same EC2 server. Every week we do some delta indexing which contains both new and updated documents. To do so we create a fresh Solr installation on a new EC2 server and attach the base index to it. We use composite id and the documents are distributed based on the group it belongs to. e.g groupId!documentId We noticed that some of the documents belonging to the same group goes to different shards, so we delete and re-index all the documents of that group to make them get into the same shard. This is happening only during delta indexing and not during the full clean indexing. This has increased our delta size and the time taken to index. Are we doing something wrong here? Can you please help us with your suggestions? Thanks, Modassar Ather
