It's not as easy as just looking at historical metrics.

What could constitute an outage could be several metrics like error rate and latency.

I'm having a hard time finding how people usually track this so you can say X service has been down X amount of time this year.

Do you just update a spreadsheet to track downtime for a service?

  • This question as it stands looks a bit vulnerable to "primarily opinon based". It is interesting though, but would lead to open-ended discussion unless some constraints were placed on the scope. – Bruce Becker Sep 10 at 15:41
  • Nagios checks status, status.io, internal cachetHQ for more sensible issues, Jira, github issues on private repos, ticketing system... That's just where I may fill up something to track an outage actually, so indeed that's very broad and probably opinion based as well. – Tensibai Sep 10 at 15:44
  • @Tensibai but does it all roll up to a spreadsheet or something? How can you use error budgets correctly if you dont keep a running tally of the time the service was down? – red888 Sep 10 at 15:48
  • Depends on which of the 96 services we're talking about, we have some "Business Intelligence" reports done from aggregation from various source, but nothing close to a "One fits them all" system to track all application/services. I don't know of a silver bullet here, people use what work for them in their environment and teams – Tensibai Sep 10 at 15:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.