Many moons ago, someone made the very common-sensical suggestion that developers be given a check-box that would tell The Scheduler to never serve a request from a cold instance. That was, as I noted, many moons ago in a thread essentially the same as this. Since then, the subject of this thread has been oft repeated. Not contributing anything I know. Hoping as a SMB user that these threads eventually end. -stevep
On Sunday, February 3, 2013 6:38:22 PM UTC-8, Michael Hermus wrote: > > Jeff's point is exactly right: many start-ups and developers subject > themselves to the lock-in and quirks of GAE because they hope to eventually > become wildly successful, and they recognize the incredibly hard challenge > of managing a rapidly scaling infrastructure with limited resources. > > However, the cold-start issue adds an unwelcome and unnecessary obstacle > to success, by subjecting a significant percentage of users to unacceptably > high latencies. There are more than enough obstacles to success already, as > most entrepreneurs are acutely aware, and so for those experiencing this it > is extremely frustrating, bordering on maddening. > > One post by a GAE Product Manager or engineer saying the following would > go a really long way with most of us: > > "We hear you.We recognize this as a significant problem. We are on it." > > On Sunday, February 3, 2013 5:30:21 PM UTC-8, Jeff Schnitzer wrote: >> >> On Sun, Feb 3, 2013 at 2:50 PM, Kaan Soral <[email protected]> wrote: >> > >> > Was your app low traffic? In my opinion there is no point in using GAE >> for a >> > non-extreme traffic app (or the possibility of extreme traffic, that's >> the >> > dream), because you restrict yourself in extreme ways rather than using >> > mysql and stuff, and chill. >> >> High-traffic apps rarely start that way. If apps are stillborn because >> performance sucks during the early-adopter and growth stage, then GAE >> is really only useful for Google and we're pretty much all wasting our >> time. >> >> Also: I have yet to see any hard evidence that high-traffic apps do >> not suffer the same cold start problem as low-traffic apps. It could >> easily be that the number of cold start requests are simply lost in >> the volume of successful requests. After all, users don't report hung >> http requests; they just reload the page and assume it was their ISP's >> fault. >> >> Gafal: Sorry to see you go, but I totally understand. This issue is >> deeply frustrating for me too. >> >> Jeff >> > -- You received this message because you are subscribed to the Google Groups "Google App Engine" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To post to this group, send email to [email protected]. Visit this group at http://groups.google.com/group/google-appengine?hl=en. For more options, visit https://groups.google.com/groups/opt_out.
