Hi Solr Users/Dev,
What needs to happen to core, shard and node recovery during startup
of solr? Currently, is not taking place in the right order so the solr goes
into an unresponsive state for a few minutes even when shard is marked
active. This causes a handful of queries to fail to return 504/timeout in
that time frame? This issue is reported SOLR-17049
<https://issues.apache.org/jira/browse/SOLR-17049>
The way I worked it around is adding a firstSearcher listener which allows
core to recover during that time period and the shard is not marked active
until firstSearcher queries are replayed but this does not work well when
the scale of cluster varies then it need to have smaller or larger set of
queries. What other workaround do you suggest?
Thank you,
Rajani
