I’m using influxdb with only 1 bucket since 1 year.
Looking on the filesystem I can identify my bucket and two other buckets. I thinks these other buckets are the system buckets _tasks and _monitoring. Looking the size of this buckets in MB, I can see this:
du -sm *
36278 MB - 8c29a7226bbf4323
1 MB - bf095049528817b4
54 MB - cc5be0f7587d660d (this is my bucket with 54 MB, without retention)
and I thinks the other two are _tasks and _monitoring. Why are they so big (using a retention of 7 and 3 days)? And I can see that they are growing in the time, some months before they are only 25 GB.
I also doing nothing special with this influxdb. No task, no telegram, nothing. Only 1 bucket saving data.
Can anyone help? How to decrease this data? How to stop growing?
Nobody can help? At least some data from other systems to be able to compare this strange behavior?
How big are your system buckets named _tasks and _monitoring?
But I solved the problem by deleting all data, that is older than 3 days. I’m not sure if I configured something wrong, or it was a bug. Hopefully it won’t get that big again.
Interesting. The system buckets have default retention periods that should delete data automatically as it expires.
_monitoring: 7 days
_tasks: 3 days
Maybe the retention enforcement service wasn’t functioning properly and expired data wasn’t getting dropped. But this is something that would’ve (should’ve) showed up in InfluxDB logs.
If all else fails, restarting influxd would also restart the retention enforcement service and hopefully clear out data that shouldn’t be hanging around.
