OH "switching to SLO-based alerting dropped our base rate of alerts by 85-90%. Way more folks are willing to be on call now." 

#deleteyourpagingalerts
-
-
So initially 5 common against all 100+ services, with moving to more service specific SLIs this year? Curious on your initial 5? RED + saturation + ? Any batch or stream services?
-
Yes. We had common 5 for all - request latency & error rate metrics; no availability/throughput metric. End user & server side response times; errors, http errors, exceptions/min. Only for our critical web apps. No streaming; batch will be taken up this year. Waiting for a tool.
Kraj razgovora
Novi razgovor -
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.