Hi!
Configuration: Zabbix 5.0.3 + TimescaleDB 1.7.4 + PG12
In General all cleaning parameters enabled (see attach)
Configured compression (and it's works)
Logs for housekeeper (tables not on TimescaleDB):
First dates on graph - system start monitoring >1200 hosts
Red arrow show moment when first most-old compressed chunks dropped
But I found that db size growth each day ~500MB. (~2 month observation period). My expectation - after first drops db size must stop growth.
Trying measure all non-timesclaled tables size ~2.5GB
Somebody have any idea where space disappeared ? Maybe some garbage after chunk drops or internal info of timescaledb plugin saved ?!
Configuration: Zabbix 5.0.3 + TimescaleDB 1.7.4 + PG12
In General all cleaning parameters enabled (see attach)
Configured compression (and it's works)
Logs for housekeeper (tables not on TimescaleDB):
Code:
3118311:20210125:215814.603 housekeeper [deleted 0 hist/trends, 0 items/triggers, 288 events, 127 problems, 36 sessions, 0 alarms, 37 audit, 0 records in 0.895198 sec, idle for 1 hour(s)] 3118311:20210125:225815.093 executing housekeeper 3118311:20210125:225815.542 housekeeper [deleted 0 hist/trends, 0 items/triggers, 304 events, 105 problems, 36 sessions, 0 alarms, 36 audit, 0 records in 0.447173 sec, idle for 1 hour(s)]
First dates on graph - system start monitoring >1200 hosts
Red arrow show moment when first most-old compressed chunks dropped
But I found that db size growth each day ~500MB. (~2 month observation period). My expectation - after first drops db size must stop growth.
Trying measure all non-timesclaled tables size ~2.5GB
Somebody have any idea where space disappeared ? Maybe some garbage after chunk drops or internal info of timescaledb plugin saved ?!
Attached Files