Improve flaky alerts
Averaged over a month and removing the "real" alerts via something like sum(sum_over_time(ALERTS{alertstate="firing",alertname!~"..."}[60d])) by(alertname)
alertname | Value | |
---|---|---|
ResourceRequestExceeded | 121290 | |
QueuesFull | 4928 | |
CPU Load | 1394 | |
Missed reboot cycle | 353 | |
IOWait | 92 |
Edited by Michael Hofmann