This message was deleted Apache Druid #troubleshooting

Join Slack

This message was deleted.

# troubleshooting

Slackbot

02/14/2023, 8:41 AM

This message was deleted.

Gian Merlino

02/14/2023, 10:17 AM

This can happen if you have mixed granularities in your various ingestion operators, and those mixed granularities overlap. In general it's not recommended to overlap mixed granularities. Meaning: this error shouldn't happen if you stick to a single granularity for a datasource. Please let us know if it does, and provide any info you can about how you got into that state!

Antonio Sarro

02/14/2023, 2:50 PM

I think the problem was setting

druid_storage_storageDirectory

when i'm using S3 buckets for segments. So i think that when i deleted some segments, inside the local folder there were some leftovers. I deleted this configuration and create a new docker containers setup, and for now it seems go flawless

Antonio Sarro

02/14/2023, 4:31 PM

Another problem that i have now is this:

DELETE /druid/coordinator/v1/datasources/{dataSourceName}/segments/{segmentId}

When i call the following query its return me 500 http status, but the segment is deleted succefully. How can i check if the invalidation is succeded with an API call?

Gian Merlino

02/14/2023, 4:45 PM

You could look at the

sys.segments

table (https://druid.apache.org/docs/latest/querying/sql-metadata-tables.html#segments-table) of Druid SQL

Antonio Sarro

02/14/2023, 4:53 PM

So is it a bug that this API return 500 and delete the segment too?

Gian Merlino

02/14/2023, 4:58 PM

I am not sure, sometimes 500 is a bug and sometimes it is due to some transient error and you can retry on the client side. If it happens repeatedly, it is probably a bug

Open in Slack

Previous Next