reduce influxdb retention period and store interval

[MarkTNO]

No description provided.

[MarkTNO]

I restarted influx on the test env to apply these settings.
The _internal db size (and memory usage) should decrease now over the span of a week: the old retention setting was 1 week.
We wait for this before deploying on prod.

[cwang39403]

I restarted influx on the test env to apply these settings. The _internal db size (and memory usage) should decrease now over the span of a week: the old retention setting was 1 week. We wait for this before deploying on prod.

Was the old retention setting of 1 week applied manually in the past?

If I run the following query on Test InfluxDB
SHOW RETENTION POLICIES ON "_internal"

I see "monitor" policy is still there. I was expecting to see "short" policy instead. Do I miss anything?

[MarkTNO]

You are right, that should have been there. Maybe it wasn't set properly or it got removed automatically.
The setup.sh script does not work properly when the influxdb contains a lot of data, since the startup of influxdb takes too long.
In the past this retention setting was set to default.
Now I updated this manually, we should keep an eye on this.
The store interval was setup properly (via influxdb.conf): since last Friday the size of the _interval database decreased from 3GB to 1.6GB.

[cwang39403]

Another reason to have a simple monitoring mechanism in OMOTES, perhaps?

Since we have now created the retention policy manually, maybe good to mention that in the README?

[MarkTNO]

Readme updated 5ade481

On the NWN test environment after 1 week (old retention period) with store interval from 10s to 5 min and retention from 7d to 1d:

• _internal db size from 3GB to 54 MB
• influxdb memory usage, after a influxdb container restart, a factor 6/7 lower
• lower cpu usage