Even with a decent CI/CD pipeline, I still find that the first 10-15 min after a production deploy are strangely manual.
the build passed, the deploy succeded, but there is still a gap where i don't really know yet whether the change is actually behaving correctly in prod.
MOST OF THE TIME THIS TUNRS INTO SOME COMBINATIION OF:
- watching dashboard
- checking recent error logs
- waiting to see if alerts fire
- deciding whether to leave it alone or investigate further
I'm curious how other teams handlet this moment.
Do you mostly rely on manual observation here, or do you have some lightweight way to verify production behavior right after deploy?
I'm interested in workflows that don't require a heavy observability stack.
0 comments
—
Ask HN: How do you automate the anxiety after a deploy · HackerTrans
Even with a decent CI/CD pipeline, I still find that the first 10-15 min after a production deploy are strangely manual.
the build passed, the deploy succeded, but there is still a gap where i don't really know yet whether the change is actually behaving correctly in prod.
MOST OF THE TIME THIS TUNRS INTO SOME COMBINATIION OF:
- watching dashboard
- checking recent error logs
- waiting to see if alerts fire
- deciding whether to leave it alone or investigate further
I'm curious how other teams handlet this moment.
Do you mostly rely on manual observation here, or do you have some lightweight way to verify production behavior right after deploy?
I'm interested in workflows that don't require a heavy observability stack.