This message was deleted.
# general
s
This message was deleted.
k
The whitelist is set using the Coordinator Dynamic properties https://druid.apache.org/docs/latest/configuration/#dynamic-configuration
j
How do I specify "all" datasources?
k
If
killDataSourceWhitelist
is not set or empty, then kill tasks can be submitted for all datasources.
j
Ah, okay, that's not documented. Hmm.. Our coordinator is not issuing kill tasks and I assumed it was because we had not set the whitelist.
k
Actually it is
j
Hmm. I see:
k
oh - like you didn’t have kill tasks turned on? Right, they are not on by default
j
It is turned on but still no kill tasks 😢
k
You have something like this set?
Copy code
druid.coordinator.kill.pendingSegments.on=true
druid.coordinator.kill.on=true
druid.coordinator.kill.period=PT4H
druid.coordinator.kill.durationToRetain=PT24H
druid.coordinator.kill.maxSegments=200000
j
I don't have
Copy code
druid.coordinator.kill.pendingSegments.on=true
However, I have just learned of
druid.coordinator.kill.bufferPeriod
I'm going to try setting this to
PT6H
and see if I get any success with that (the default appears to be 30 days) and I have retention at 7 days.
Config that wasn't working (kill tasks not running):
Copy code
$  cat runtime.properties |grep druid.coordinator.kill
druid.coordinator.kill.on=true
druid.coordinator.kill.period=PT1H
druid.coordinator.kill.durationToRetain=P1D
druid.coordinator.kill.maxSegments=10000
Config that is working (kill tasks being issued hourly):
Copy code
$  cat runtime.properties |grep druid.coordinator.kill
druid.coordinator.kill.on=true
druid.coordinator.kill.period=PT1H
druid.coordinator.kill.durationToRetain=P1D
druid.coordinator.kill.bufferPeriod=PT6H
druid.coordinator.kill.maxSegments=10000
Very first kill interval:
Copy code
"interval": "2023-12-30T22:00:00.000Z/2024-02-07T15:00:00.000Z",
Strange...
@Kyle Ahn I have come back to check after applying these settings and I can see that Deep Storage is still not being cleaned up. I believe we have an issue with version 28.0.1 not cleaning Deep Storage. I did not have an issue before upgrading.
The introduction of
druid.coordinator.kill.bufferPeriod
seems to have broken some things.