Here are some potential routes for debugging / profiling in general: - Make use of the Cloud Debugger <https://cloud.google.com/debugger/docs/setting-up-java-on-app-engine> to inspect variables - Make use of AppStats <https://cloud.google.com/appengine/docs/java/tools/appstats> to profile RPC call timings - Make use of custom timing code before and after relevant code blocks/lines as an alternative to AppStats - Make use of Java memory use inspection calls before and after relevant code blocks/lines to track memory consumption - Deploy the app to a Flexible Environment Custom Runtime <https://cloud.google.com/appengine/docs/flexible/custom-runtimes/> app using a -compat runtime <https://cloud.google.com/appengine/docs/flexible/custom-runtimes/build#configuring_the_dockerfile> for Java (so that you existing GAE code shouldn't need to change), using the Dockerfile to run commands during container building <https://cloud.google.com/appengine/docs/flexible/custom-runtimes/build> which install and configure a more comprehensive Java memory profiler (of your choice). You could then swap out the default serving version for this version long enough to observe the memory use on requests, and its causes.
I hope these methods are useful. The last two are most relevant to out-of-memory errors. Let me know if you have any more questions about implementing any of them - I'll be happy to help! Cheers, Nick Cloud Platform Community Support On Monday, September 12, 2016 at 5:58:07 AM UTC-4, Ice13ill wrote: > > Hi Nick, thank you for your response. > It seams that in time, many request have begun to need more and more > memory, and some requests that need more processing time (they take longer > and need new instances). I believe there are some memory problems, but I am > not sure because many requests simply fail with code 104 and i see no out > of memory exceptions. > An advice regarding memory profiling for app engine / jetty web apps (if > only locally available) would be much appreciated! I also posted on > stackoverflow here: > http://stackoverflow.com/questions/39447618/google-app-engine-java-memory-profiling-on-local-dev-server > > Thank you! > > On Friday, September 9, 2016 at 2:36:04 AM UTC+3, Nick (Cloud Platform > Support) wrote: >> >> Hey Folks, >> >> Have you taken a look at the documentation >> <https://cloud.google.com/appengine/articles/deadlineexceedederrors> >> related to Deadline Errors to determine what might be happening in your >> apps? Do your request response times generally hover around the 60 second >> limit, or is it conceivable that they could inflate above that limit with a >> degradation of any service they rely on synchronously? >> >> Cheers, >> >> Nick >> Cloud Platform Community Support >> >> On Tuesday, September 6, 2016 at 2:59:47 PM UTC-4, Mauricio Lumbreras >> wrote: >>> >>> Hello >>> are you suffering this in a specific time frame or scattered during the >>> whole day? >>> I suffered some issue with version 1.9.44 and we were downgraded to >>> 1.9.43 >>> We face some problems from 31th aug >>> Regards >>> Mauricio >>> >>> >> -- 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 https://groups.google.com/group/google-appengine. To view this discussion on the web visit https://groups.google.com/d/msgid/google-appengine/3cbf99bc-19b7-4f83-aae1-269e27ca82c2%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
