Thanks to those who responded. My observation that the blips occur at apparently random times during the handler run lead me to be a little suspicious of the new interpreter theory - I only do fairly minor "from xxx import yyy" imports in some of the sections that have included blips.
I'll add log points around my imports, and that will show if they really are the culprit. I'll post back here when I get more data. If it does turn out to be this, then I'd suggest it would be fairer for app engine to take the cpu cost of loading new interpreters out of the calculation, as they do with datastore and other API calls. Cheers! Greg. On Dec 10, 9:47 am, "Sharp-Developer.Net" <[EMAIL PROTECTED]> wrote: > I see the same picture described by me here: > > http://groups.google.com/group/google-appengine/browse_thread/thread/... > > Marzia says it could be becouse of python modules load by new instance > of interpreter. > -- > Alexhttp://sharp-developer.net/ > > On Dec 9, 4:34 am, Greg <[EMAIL PROTECTED]> wrote: > > > I uploaded my first real app yesterday, and have had seventy people > > hit it. Everything seems to be working like a charm, but those nasty > > little average CPU warnings got me worried. My app uses memcache > > aggressively, does very few datastore writes, and no page does > > extensive processing, so I was surprised to see any warnings. > > According to a post I found, I can have two requests of over 1000Mc a > > minute, before Appengine starts raising quota exceptions. It also said > > that the dashboard figure included datastore megacycles which aren't > > included in the quota calculations, and that the log gives accurate > > calculations. > > > So I did some analysis of the logs. Out of more than 700 non-static > > requests, there were 9 high CPU warnings. The URLs that triggered > > warnings were pretty diverse, including some that do very little work, > > and other hard working pages haven't figured yet. Four were for the > > same URL, which logged four events, and in each case there was a long > > (0.5 to 1sec) delay between two of the events - but two different > > events each time. In one case it was after the request has started but > > before the first event, and in another it was between the second and > > third events, etc. I.e. the delay was not always at the same part of > > the code - it wasn't datastore or memcache calls that caused it. > > > I don't have enough data to be completely certain, but it seems that > > there are fairly random 0.5 to 1 second "blips" that occasionally hit > > my application, and cause CPU warnings. I could imagine these are > > caused by the OS or other applications, and my concern is that my > > application could get quota exceptions that are not caused by external > > factors. And when billing arrives, I'll get charged for them instead. > > > Questions: > > 1. Has anyone else seen this behaviour? Is everyone getting occasional > > CPU warnings, and is anyone getting actual quota exceptions because of > > them? > > > 2. Has Google any idea what is causing the blips? > > > Apart from this, I've enjoyed app engine - once little niggles like > > this are ironed out, it is going to be a great platform. Can't wait > > for it to come out of preview. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "Google App Engine" group. To post to this group, send email to google-appengine@googlegroups.com 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 -~----------~----~----~----~------~----~------~--~---
