"Enforced Rate" in task queues ruining performance

324 views
Skip to first unread message

Brian Olson

unread,
Aug 1, 2011, 5:02:54 PM8/1/11
to google-a...@googlegroups.com
Why do we sometimes get vastly underperforming task queues?
I've attached a screenshot showing how we're getting 0.62/s on a task queue that's configured to run at 10/s. (and another getting 1.25/s where it should be 10/s).
What's the magic word to tell AppEngine "yes, really, run at full speed"?
Screen shot 2011-08-01 at 4.58.59 PM.png

Robert Kluin

unread,
Aug 2, 2011, 12:06:09 AM8/2/11
to google-a...@googlegroups.com
Excellent question. Hopefully a Googler will give a detailed
explanation of all the causes.


What is your average latency within the slow queues? What about the
error rates? If I'm not mistaken, those two things have been used to
throttle queues at various times.

Robert

> --
> You received this message because you are subscribed to the Google Groups
> "Google App Engine" group.
> To view this discussion on the web visit
> https://groups.google.com/d/msg/google-appengine/-/gXr_OuW6oTMJ.
> To post to this group, send email to google-a...@googlegroups.com.
> To unsubscribe from this group, send email to
> google-appengi...@googlegroups.com.
> For more options, visit this group at
> http://groups.google.com/group/google-appengine?hl=en.
>

Dan Lines

unread,
Aug 2, 2011, 9:08:35 AM8/2/11
to google-a...@googlegroups.com
I am having the same problem.  Today has been the worst performance so far with the enforced rate down at 0.16/s.  The application is basically non functional at this rate....

Could we please get some information on this.

- Dan
Screen shot 2011-08-02 at 9.06.37 AM.png

Johan Euphrosine

unread,
Aug 2, 2011, 10:08:41 AM8/2/11
to google-a...@googlegroups.com
Can you both share your appid, so we can investigate deeper ?

Also feel free to fill a production issue about it if this is directly
affecting your operation:
http://code.google.com/p/googleappengine/issues/entry?template=Production%20issue

> --
> You received this message because you are subscribed to the Google Groups
> "Google App Engine" group.
> To view this discussion on the web visit

> https://groups.google.com/d/msg/google-appengine/-/6cczXBYIK6IJ.


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

--
Johan Euphrosine (proppy)
Developer Programs Engineer
Google Developer Relations

Dan Lines

unread,
Aug 2, 2011, 3:47:57 PM8/2/11
to google-a...@googlegroups.com
Johan,

I have submitted a production issue for this problem.  

If you have any questions or need to contact me feel free to email me directly.

Thank you,

- Dan

Jan Zawadzki / Hapara

unread,
Sep 2, 2011, 2:10:44 PM9/2/11
to Google App Engine
Same here Johan - I think this is system-wide.

It is definitely tied to the new application settings, in particular
the min. pending latency.

This sort of stuff seriously needs to be documented better folks - the
ones affected by it the most are people using App Engine in earnest;
for us and possibly for others this is a system-down sort of event.


Jan

On Aug 3, 2:08 am, Johan Euphrosine <pro...@google.com> wrote:
> Can you both share your appid, so we can investigate deeper ?
>
> Also feel free to fill a production issue about it if this is directly
> affecting your operation:http://code.google.com/p/googleappengine/issues/entry?template=Produc...
>
>
>
>
>
>
>
>
>
> On Tue, Aug 2, 2011 at 3:08 PM, Dan Lines <dli...@aprigo.com> wrote:
> > I am having the same problem.  Today has been the worst performance so far
> > with theenforcedratedown at 0.16/s.  The application is basically non

stevep

unread,
Sep 2, 2011, 10:20:00 PM9/2/11
to Google App Engine
Having commented many times about how task queues need to be improved,
I guess I can only +1 this.

Not sure what the deal is with TQs. They offer so much power, yet seem
an afterthought.

Maybe the guy/gal that developed them moved to another project, and no
one wants this "step child" process?
Reply all
Reply to author
Forward
0 new messages