I delete the branch about #6283 by mistake. So I can't modify that pull request.
There are some comments. 1. Mysql default transaction isolation is REPEATABLE_READ, treat it as READ_COMMITTED will reduce insert id conflict. 2. My segments table is too big to select the overlaps segment items. Add an index to 'dataSource used end' is work well for the most of scenarios. It will speed up getting recently segments for select and insert. Add an index to 'dataSource used start' is a good idea, but I didn't include it in this patch. 3. I think 'select and insert' doesn't need within transaction. (Fix same issue with phase 1) 4. sync's performance is better than concurrent retry in my business. [ Full content available at: https://github.com/apache/incubator-druid/pull/6336 ] This message was relayed via gitbox.apache.org for [email protected]
