[google-appengine] Re: appstats analysis - Same request - Too much difference in serving time.

Hi, I have posted this post, with attachments twice. But it did not make into the group. Any limitation on posting with attachments. Any settings to be enabled. I have problem with serving the same request with big difference in the datastore and memcache timings. Please see the below email. O

[google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Hi Rajesh, Results like these are generally to be expected (at least in terms of the variance of latencies). The trade-off you make with Cloud services is that while you get the scalability and pwoer of a data-center-wide RAM cache or schemaless Datastore, you also get the variability of such

[google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Dear Nick I have noticed that this latency difference happens for a request with a new instance. The same request will have considerable less latency for the second time for the same instance. I will also consider redoing some code and do some batch calls as suggested by you. But,I have mad

[google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Hi Rajesh, The first request vs. second request differences can be explained by taking a look at the functioning of Objectify, which it appears you're using. In Objectify, the library maintains a small in-memory cache which allows it to serve requests and update entities with less latency. This

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time.

On Mon, Aug 3, 2015 at 4:43 AM, Rajesh Gupta < rajesh.gu...@veersoftsolutions.com> wrote: > Hi, > I have posted this post, with attachments twice. But it did not make into > the group. Any limitation on posting with attachments. > Sorry Rajesh, it was a temporary glitch with the group -- I have

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time.

Rajesh - if you're performing *exactly* the same execution and are simply asking about the variability - that's just sort of the nature of the distributed RPCs. There can sometimes be pretty wide range of response times in RPCs. Depending on your frameworks, there are ways to execute queries in

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Dear Nick I have noticed that this latency difference happens for a request with a new instance. The same will have considerable less latency for the second time for the same instance. Since, with java appengine, the instances are always spinning (this is a different issue and discussed in sever

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

This is not quite accurate. Objectify's session cache spans only a single request; there is no instance cache shared across requests (other than memcache). However, if that first request includes the one-time entity class registration that Objectify requires, that could easily explain extra time s

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Thanks Jeff and Nick. The entities are register in the ServletContextListener, and hence the first request and second request are the same. On Tue, Aug 11, 2015 at 6:40 AM, Jeff Schnitzer wrote: > This is not quite accurate. Objectify's session cache spans only a single > request; there is no

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Hi Rajesh, Do you think it might be possible for you to duplicate your app, cut out unnecessary functionality, and post this to a public issue tracker issue from which we could take a deeper look at this and possibly determine why it's beh

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Hello Nick, It is difficult to cut out the app and post the code on the issue tracker. I can put lot of logs in the code. I will give the details of the app-id, and will show you the logs. You can also see the appstats for the same requests. The datastore get() and query() take longer time. If

Re: [google-appengine] Re: appstats analysis - Same request - Too much difference in serving time

Hey Rajesh, Whatever you do when you post your issue tracker issue, make sure that you include enough information that can allow somebody to either observe the issue on your app, or reproduce the issue independently, and be sure to respond to any replies on the thread asking for extra needed in