Re: [google-appengine] QPS versus Req/sec

527 views
Skip to first unread message
Message has been deleted

Nick Johnson (Google)

unread,
Jan 27, 2011, 11:23:21 PM1/27/11
to google-a...@googlegroups.com
Hi smwatch,

On Fri, Jan 28, 2011 at 12:30 PM, smwatch <show...@gmail.com> wrote:
Can some Google Expert please clarify this for us?

What is the difference between QPS versus Req/sec

These are the same thing - though QPS (Queries Per Second) is a slightly broader term.
 
When will google api limits start hitting, we read it is 400 Requests per Secs

is the limit for 400 Average QPS or 400 Req/Sec as shown below for a given URL?

It used to be the case that if you were going to exceed 400QPS, you would need to contact us to get an exemption so you could exceed that amount without issues. That's no longer the case - your (billing enabled) app can now handle as many QPS as you like without intervention from us.

-Nick Johnson
 

Below are some of our details
Number of Instances - Details Average QPS Average Latency Average Memory
30 total 2.478 23.7 ms 10.1 MBytes

Req/Sec current Requests last 17 hrs Avg CPU (API) last hr % CPU last 17 hrs
/myauoURL
78.6 4.56M
/myurl2                 1.3

--
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-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.



--
Nick Johnson, Developer Programs Engineer, App Engine
Google Ireland Ltd. :: Registered in Dublin, Ireland, Registration Number: 368047

Albert

unread,
Jan 28, 2011, 3:50:31 AM1/28/11
to Google App Engine

> It used to be the case that if you were going to exceed 400QPS, you would
> need to contact us to get an exemption so you could exceed that amount
> without issues. That's no longer the case - your (billing enabled) app can
> now handle as many QPS as you like without intervention from us.

Hi!

If that's the case, then why are messages like "Outgrowing the
maximums? Request an increase." all over the official Docs
http://code.google.com/appengine/docs/quotas.html#Billable_Quotas_and_Fixed_Quotas
?

Are the docs outdated, or are you referring to something else here?

Thanks!



Albert

Robert Kluin

unread,
Jan 28, 2011, 12:51:05 PM1/28/11
to google-a...@googlegroups.com
I suspect it is because there are lots of different quotas you might
bump into depending on your app. Some of them are not customizable in
the billing settings.


Robert

Ikai Lan (Google)

unread,
Jan 28, 2011, 1:34:49 PM1/28/11
to google-a...@googlegroups.com
Yep, Robert is correct. There are maximums such as Memcache API usage, datastore usage, etc. The limits are set pretty high by default for billing enabled apps, but many of our heaviest applications have exceeded these limits and require quota increases.

--
Ikai Lan 
Developer Programs Engineer, Google App Engine

Brandon Wirtz

unread,
Jan 28, 2011, 1:45:44 PM1/28/11
to google-a...@googlegroups.com

If you missed the Hint in there…. If you think you are hitting quota’s consider enabling billing.   (this has helped me a few times)  (It also misled me once because I did something stupid and the quotas were trying to protect me from shooting myself in the foot, and all I did was get a bigger gun)

Reply all
Reply to author
Forward
0 new messages