Thanks Michael, that's exactly what I was looking for.

-David

On Mar 3, 5:20 am, Michael Wesner <mike.wes...@webfilings.com> wrote:
> The math is:
>
> (1000 ms/second / 75 ms/request) * 30 = 400 requests/second
>
> Google assumes that you can run all your requests in 75ms, which gives you 
> the 400req/sec number.  They say you should be able to do 30 simultaneous 
> requests.  We have been testing this lately and can only get 10, but that is 
> an ongoing discussion.
>
> Interesting though, since one query takes ~500ms according to todays status 
> page.  Hard to do 75ms on a heavier application/request. :(
>
> http://code.google.com/status/appengine/detail/datastore/2010/03/02#a...
>
> We are continuing to tune and reduce request length as always, but even with 
> fairly low amounts of traffic we hit this limit.
>
> On Mar 2, 2010, at 11:07 AM, Marc Provost wrote:
>
> > Hi David,
>
> > I don't have a precise answer, but I think you are going in the right
> > direction. The idea is to minimize the response time of your most
> > popular requests using memcache. Try to cache the html pages derived
> > from the datastore queries. It is easy to drop the request time to a
> > few hundred ms or even less that way (depending how much you can cache
> > and if you use java or python). This approach works well if your users
> > are interested in the same entities. For example a news site can
> > easily cache its more popular articles that way. For entities/queries
> > that are specific to each user, but might be reused in many pages,
> > cache it as soon as possible!
>
> > Hope it helps,
> > Marc
>
> > On Mar 2, 3:21 am, Waleed Abdulla <wal...@ninua.com> wrote:
> >> I got the same errors today on my dev app, which I'm the only user of. So 
> >> it
> >> doesn't seem to be related to how much load the app has!! I've been 
> >> noticing
> >> them on my production app as well on and off.
>
> >> Waleed
>
> >> On Mon, Mar 1, 2010 at 7:22 PM, Satoshi <satoshi.nakaj...@gmail.com> wrote:
> >>> I've got the same warnings several time today too. The peak access
> >>> rate was only 3.00 requests/sec, and the CPU time usage over the last
> >>> 24 hours is 6% (1.08 CPU hours) out of 18.50 CPU hours (I am a paying
> >>> customer).
>
> >>> Satoshi
>
> >>> On Mar 1, 6:51 pm, David <dscri...@gmail.com> wrote:
> >>>> I am losing sleep over this, so any help would be greatly appreciated!
>
> >>>> APP ID: conit-app01
>
> >>>> Since our app released about a week ago, it has been getting an
> >>>> average of about 60 requests/second.  On February 27, our app suddenly
> >>>> crashed and was down for several hours, with thousands of these errors
> >>>> appearing in the logs:
>
> >>>> Request was aborted after waiting too long to attempt to service your
> >>>> request. Most likely, this indicates that you have reached your
> >>>> simultaneous dynamic request limit. This is almost always due to
> >>>> excessively high latency in your app. Please seehttp://
> >>> code.google.com/appengine/docs/quotas.htmlfor more details.
>
> >>>> Since getting this error, I filled out a request to increase this
> >>>> limit at:
> >>>http://code.google.com/support/bin/request.py?contact_type=AppEngineC...
>
> >>>> This request was denied, because, "your app has been using, over the
> >>>> past 24 hours, on average 60 QPS with a peak of ~135 QPS; thus you're
> >>>> well under the 500 QPS limit described above."
>
> >>>> Since this crash, I've also been working to decrease calls to the
> >>>> datastore, and I think our average CPU time has decreased around 30%.
> >>>> In the dashboard, one of our pages still appears "yellow" under the
> >>>> column "Average CPU (API)", with a speed of about 1100.  This page is
> >>>> about 6% of the volume of our app.  The other pages don't have any
> >>>> warnings.  We are well within the limits of our billing.
>
> >>>> I would feel much better if I could understand the math/metrics that
> >>>> go into producing this error, so it doesn't happen again.  How can I
> >>>> know if my page request times are low enough?  If I add a new page
> >>>> with a higher CPU time, how can I know if it would make the app crash?
>
> >>>> Any help or references to details on this error would be appreciated.
>
> >>>> Thank you in advance.
> >>>> -David
>
> >>> --
> >>> 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-appeng...@googlegroups.com.
> >>> To unsubscribe from this group, send email to
> >>> google-appengine+unsubscr...@googlegroups.com<google-appengine%2bunsubscr...@googlegroups.com>
> >>> .
> >>> For more options, visit this group at
> >>>http://groups.google.com/group/google-appengine?hl=en.
>
> > --
> > 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-appeng...@googlegroups.com.
> > To unsubscribe from this group, send email to 
> > google-appengine+unsubscr...@googlegroups.com.
> > For more options, visit this group 
> > athttp://groups.google.com/group/google-appengine?hl=en.

-- 
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-appeng...@googlegroups.com.
To unsubscribe from this group, send email to 
google-appengine+unsubscr...@googlegroups.com.
For more options, visit this group at 
http://groups.google.com/group/google-appengine?hl=en.

Reply via email to