Hi Robert, there is a undocumented feature called edge caching. If your handler returns the Cache-Control header a local google datacenter will create a copy of the output and serve it to clients directly. If a page is edge cached you don't consume any cpu quota and only have to pay for traffic. Maybe you can make use of that feature for your ticker, I don't think that the usability would be harmed if you let google cache the ticker for 10-20 seconds.
Regards Philip On Aug 26, 10:07 am, hovu96 <[email protected]> wrote: > I post this to the groups because the Google App Engine Support is neither > answering my questions I posted to the request form > (http://code.google.com/support/bin/request.py?contact_type=AppEngineC...) > nor increasing the "CPU-minutes/minute" limit. > > A few weeks ago (during the last Apple keynote) we redirected all users from > our german news website (www.mactechnews.de) to the app engine site > (mtn-live.de/ticker/). We expected about 1000 request per second but in the > first 5 minutes we received 1500 request per second. Before this event we > increased the billing dollars but we weren't aware of the "72 > CPU-minutes/minute" limit. It is a very simple website that just fetches > some entries from the db and renders them to the html. All other resources > cached "public". > > Here are some screenshots that shows the situation: > > CPU Seconds / Second: > > <http://my-upload.appspot.com/serve/cpusecspersec.png?key=AMIfv96C4JpG...> > > CPU Milliseconds / Second: > > <http://my-upload.appspot.com/serve/millisecspersec.png?key=AMIfv96h8j...> > > Errors / Second: > > <http://my-upload.appspot.com/serve/errorspersec.png?key=AMIfv95IspnDt...> > > Instances: > > <http://my-upload.appspot.com/serve/instances.png?key=AMIfv95j7esCZrUN...> > > Requests / > Second:http://my-upload.appspot.com/serve/requestspersec.png?key=AMIfv97QiWw... > > We think the problem is shown in "cpusecspersec.png": The number of CPU > seconds per second is larger then 72 (it is about 100). > > After so many 500 errors we decided to redirect to our old ticker server > (which is hosted on windows 2008). Some requests contain this dashboard logs > message: > Request was aborted after waiting too long to attempt to service your > request. This may happen sporadically when the App Engine serving > cluster is under unexpectedly high or uneven load. If you see this > message frequently, please contact the App Engine team. > > But we really want to use App Engine for that and are planning to implement > more feature like chatting on that page or some other community features... > but this isn't possible at all if even the simplistic basic page can't be > handled properly. > > So our questions are: > > 1) Could you please increase the per minute quota ("CPU-minutes/minute") for > the app id "sy-mtn" ? (The think the next Apple keynote will take place in > September so it would be great to increase the limit as soon as possible) > 2) Do you see any other (per minute) quota that we could reach? > > Thanks, > Robert -- 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.
