Christina wrote:

> *We have resolved the first issue by fixing the underlying bug.* 

I think one good option for longer-term stability is one we've identified 
before - we need a "stable" release channel where Google won't alter the 
environment until it's thoroughly tested, meaning deployed to those who 
aren't on stable and run without error for a week/month.

Also, if I were you at some point I'd start to deploy on weekends (which I 
assume are quieter, and fewer businesses are affected by downtime).  I 
understand you don't want to have the whole full-stack engineering team in 
on weekends, but deploying in the middle of the week is insane.  Have some 
on-call, pay them more, just don't think adding one more test will fix this 
permanently.  Not sure when GAE is important enough to Google to make this 
work, but doing anything less just raises the chance that you're taking 
businesses offline.

Any chance you could tell us when you're about to alter the stack, similar 
to how you warn about the M/S downtime?  That way we can be around and 
pre-warn our own customers, just in case.

One class of problem that won't go away easily are non-GAE changes.  For 
example, someone in a far-flung corner of Google deciding that they must 
block Cloudflare-like requests for all Google properties. No idea how to 
fix that.

Richard

-- 
You received this message because you are subscribed to the Google Groups 
"Google App Engine" group.
To view this discussion on the web visit 
https://groups.google.com/d/msg/google-appengine/-/aeVysGiNoVEJ.
To post to this group, send email to [email protected].
To unsubscribe from this group, send email to 
[email protected].
For more options, visit this group at 
http://groups.google.com/group/google-appengine?hl=en.

Reply via email to