Apache Druid

Seems like that the queries are timing out. I think the default timeout is 5 minutes. A fix for the same is to set the query timeout value to a larger number. Check the following configuration properties:
1. `druid.server.http.defaultQueryTimeout`
2. `druid.server.http.maxQueryTimeout`
And the query context parameter `timeout`.

Regarding the metadata refresh, how often do you see this message. Can you also attach the cause for the refresh failure (it should be present as additional logs). It shouldn’t be detrimental to the performance if it occurs sporadically.

Hi <@U031AD802MN> thanks for taking a look. I can only find below error message in druid-broker logs related to metadata refresh:

```2023-02-07T21:00:52,589 WARN [DruidSchema-Cache-0] org.apache.druid.sql.calcite.schema.SegmentMetadataCache - Metadata refresh failed, trying again soon.```
 I believe the stack-trace here in the above message is related to `SegmentMetadataCache` which points to metadata refresh ?

I see this happening almost every minute, is it something that broker is not able to get the segment metadata from mysql ?