This message was deleted Apache Druid #troubleshooting

Join Slack

This message was deleted.

# troubleshooting

Slackbot

01/19/2023, 12:51 PM

This message was deleted.

Brandon Tan

01/20/2023, 4:17 AM

1. Timeouts: Please refer to https://druid.apache.org/docs/latest/querying/lookups.html#configuration

Ben Krug

01/20/2023, 9:14 PM

I don't know of any metrics, but in my experience, URI-based lookups do better than jdbc-based, in case that helps.

✅ 1

Michael Taranov

01/22/2023, 8:19 AM

Thank @Brandon Tan , what about retry policy? how it handled ? Any metrics fired?

Brandon Tan

01/23/2023, 2:13 AM

Let me dig into this. If it’s not in the docs, it does not exist.

Brandon Tan

01/23/2023, 3:25 AM

How about these : https://druid.apache.org/docs/latest/querying/lookups.html#saving-configuration-across-restarts 1.

druid.lookup.coordinatorFetchRetries

druid.lookup.lookupStartRetries

Brandon Tan

01/23/2023, 4:36 AM

Some more info here: When you update the lookup, the following happens: 1) The coordinator updates the lookup in the DB 2) The coordinator periodically syncs the latest version of each lookup to all nodes 3) Alternatively, a node may choose to sync to the coordinator on start-up itself We can control the period with which the coordinator syncs the latest version of the lookups to all nodes by setting

druid.manager.lookups.period

. The default period is 2 mins. Kindly set

druid.manager.lookups.period

to 30000 ms (30 Sec) to update the lookups bit faster.

✅ 1

Michael Taranov

01/24/2023, 7:52 AM

Thanks @Brandon Tan !

Open in Slack

Previous Next