asynchronous queries and updates in SOLR?

[jnioche]

Maybe https://github.com/inoio/solrs would be useful?

[jnioche]

see https://solr.apache.org/guide/solr/latest/deployment-guide/solrj.html

ConcurrentUpdateHttp2SolrClient - just like ConcurrentUpdateSolrClient but using Http2SolrClient instead. This class is experimental therefore its API’s might change or be removed in minor versions of SolrJ.

Not sure whether CloudSolrClient is asynchronous

[CloudSolrClient] (https://solr.apache.org/docs/9_0_0/solrj/org/apache/solr/client/solrj/impl/CloudSolrClient.html) - geared towards communicating with SolrCloud deployments. Uses already-recorded ZooKeeper state to discover and route requests to healthy Solr nodes.

[mvolikas]

My understanding is the following:

• ConcurrentUpdateHttp2SolrClient is not experimental in version 9.7.0 and seems promising for the StatusUpdater and Inderer bolts where it can batch updates without changing the way we submit documents.

• Regarding querying in the spouts we could use the requestAsync of the Http2SolrClient with futures.

• I'm not sure about the benefits of the CloudSolrClient, to be honest. The API seems to be synchronous from the caller's perspective. Additionally, since we currently have a single node (with multiple shards for the status collection) we will not use any of the automatic routing (e.g. load balancing requests across nodes) and failover (e.g. a node goes down) features.

Any thoughts?