Robert's tuning has done very well for us in various situations. Running a large diverse app takes a balance of these settings. When you factor in task queue performance and such, it can be a little weird. The rules are going to be different for every app, so I don't think you can have a magic formula.
It really comes down to a balance of what you want to pay vs the risk you want to take of a spike causing issues for customers. -Mike On Feb 7, 12:55 am, "Brandon Wirtz" <[email protected]> wrote: > > Did you factor into account app startup times? Some apps can easily take > > several seconds to get fired up, so they do better with a larger buffer. > For > > light apps I agree, I've found 1 or 2 is sufficient. > > Yeah, I said warm-up not startup, but same thing. > > I like lower numbers too, but if your max latency is at 1/3 your average you > will pretty much be guaranteed to have an instance per simultaneous request. > My average request is 300ish MS so I can pull off 500-600ms, but someone > with 2.5s average would spin up a lot of extra instances everytime a user > hit a page making 8 requests. And you'd eat the start up time on all those > instances, so too low of a Max can make things worse. Again check the > archive I talked about this too. -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. 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.
