I have also installed this lib and am having fun with it. personally i 'feel' the overhead is really insignificant
refer to your diagram, i'm not sure whether the blue bars are actually showing time without the profiler, i think it may mean something else (i also wanna know what the blue bar and red rectangle refer to, and will dig into the doc to find out what it actually means. anyone know?). the overhead should be the time labeled with 'overhead' (in your case, which is 0ms) my personal experience is, for most calls, the 'overhead' is always <1% of the 'real' - eric 2010/1/20 Danny Tuppeny <da...@tuppeny.com>: > I haven't used this in production yet (I only read that you can), though I'd > avoid leaving it on for long periods if you don't need it. > When running in production, it will show two bars - the one showing how much > time was taken without the profiler, and one showing how long was actually > taking (including the profiler overhead, since that's what you'll be charged > with). There's a screenshot here: > http://2.bp.blogspot.com/_DkVkyz39PB0/SyLYRJPlsgI/AAAAAAAADeg/5FiuJi3SpjQ/s1600-h/スクリーンショット（2009-12-12+8.39.08）.png > The overhead isn't insignificant, so you probably don't want to leave the > profiler turned on, but the blue bars show the time without the profiler, so > the results are still accurate :-) > Danny > > 2010/1/20 Eric Ka Ka Ng <ngk...@gmail.com> >> >> wow!! this is THE TOOL that I'm looking for long time!! previously i >> have to write codes to do similar profiling and analysis with much >> limited scales and functions. Now, THIS provides almost everything I >> want already without any efforts from developers! I could say it is >> one of the best tool on GAE. Great Work!! >> >> how about the overhead? from the doc I know we already can read from >> the INFO log about its overhead for recording the RPC events and >> storing it. for normal applications, what's your experiences or >> expectations on the added overhead? (like <1% CPU time increase?) if >> the overhead is relatively small, I think we can include it even in >> the production system without noticeable performance impact. >> >> >> many thanks, >> eric >> >> 2010/1/19 mani doraisamy <mdorais...@orangescape.com>: >> > I guess the reason why this is so damn good is because it gives an >> > integrated picture across datastore, memcache and web request, without >> > much overhead! Given that the variations are fixed (GAE's limitation >> > of not allowing all libraries/api/rpc calls), this gives the full >> > picture of the application. >> > >> > On Jan 19, 4:18 pm, mani doraisamy <mdorais...@orangescape.com> wrote: >> >> Thanks for the link, Danny. This is the best profiling tool i have >> >> used. Well done, Guido! >> >> >> >> Will get back with the performance data for off-peak Vs peak hours. >> >> >> >> thanks, >> >> mani >> >> >> >> On Jan 17, 9:37 pm, Danny Tuppeny <da...@tuppeny.com> wrote: >> >> >> >> >> >> >> >> > Have you identified what part of your page is taking additional time? >> >> >> >> > Guido van Rossum wrote a library called Appstats to help profile time >> >> > spent on API calls. I'd recommend setting it up (it works in >> >> > production) to see if you can identify where this time is being >> >> > spent. >> >> >> >> > I wrote a short article about Appstats on my blog with some >> >> > screenshots showing the sort of data you can get out of it: >> >> >> >> >> >> > >http://blog.dantup.com/2010/01/profiling-google-app-engine-with-appstats >> >> >> >> > Hope this helps. >> >> >> >> > Danny >> >> >> >> > On Jan 16, 8:24 pm, mani doraisamy <mdorais...@orangescape.com> >> >> > wrote: >> >> >> >> > > Recently, we have been running into frequent performance problems, >> >> > > especially between 7-10pm IST. Requests that used to take 600ms are >> >> > > taking almost 4 secs. >> >> >> >> > > - Has anyone faced similar problems recently? >> >> > > - What is the limit for memcache per account? How do we find the >> >> > > cache >> >> > > expiry pattern? (actual expiry Vs specified expiry) >> >> >> >> > > thanks, >> >> > > mani >> > >> > -- >> > 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. >> > >> > >> > >> > >> >> -- >> 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. >> >> > > > -- > 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. > > -- 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.
