Colin
@colin
I migrated the /paragraph observability stack from a combination of Sentry + Grafana/Prometheus + GCP Stackdriver, to Datadog. It's been a pleasure so far: - Can easily view errors / logs / traces / profiles in one place. Makes debugging substantially easier - Distributed tracing worked out-of-the-box This is in contrast to the previous stack: - Struggled a lot with Sentry (eg, distributed tracing never worked for us; challenging to reduce frontend error noise) - GCP Stackdriver logging express middleware felt hacky - Challenging to debug given multiple tools: had to consult Grafana for utilization, Stackdriver for logs, Sentry for errors, etc
2 replies
2 recasts
19 reactions
Giuliano Giacaglia
@giu
Datadog is really good!
0 reply
0 recast
1 reaction
