2 ms·
> Misconfiguration, hardware failure, power outage? Heck, maybe the DB is not actually down but there was a network failure that made it unreachable from the mo
by falcolas 12y ago
> Misconfiguration, hardware failure, power outage? Heck, maybe the DB is not actually down but there was a network failure that made it unreachable from the monitoring server.
Sorry, I should have been more specific. DB daemon being down is a cause, and should be monitored. Hardware down is a cause, and should be monitored. Network availability is a cause and should be monitored. Power outage... you get my drift.
I think the root cause of my disagreement with this document is the lack of a proper dependency tree in the alerting tool. Their tool appears to want to alert on any and every monitored problem, which necessitates limiting what you monitor for fear of a pager flood. A proper tool can address this problem correctly.
- rhizome 12y ago911: PING TO GOOGLE >100ms