Excessively high latency

[mscwd01]

Hey,

For the last week or so I have been getting a high error rate (approx
40%) on my Java app along with this error:

Request was aborted after waiting too long to attempt to service your
request because you have reached your simultaneous dynamic request
limit. This is almost always due to excessively high latency in your
app. Please see http://code.google.com/appengine/kb/general.html#activerequests
for more details.

Looking at my logs my average latency is 10,000ms.

Is anyone else having this problem? It's a real problem as I am seeing
a 500 being thrown on every other request!

Thanks

[mscwd01]

Just to add to my first message, I took a look at AppStats and
realised its PUT requests to the datastore which are driving up my
request time. Here are the stats for a typical request:

@1ms memcache.Get real=9ms api=0ms
@10ms memcache.Set real=10ms api=0ms
@21ms datastore_v3.Get real=13ms api=0ms
@36ms datastore_v3.Put real=341ms api=0ms
@388ms datastore_v3.RunQuery real=146ms api=0ms
@536ms datastore_v3.RunQuery real=16ms api=0ms
@552ms datastore_v3.RunQuery real=21ms api=0ms
@595ms datastore_v3.Put real=338ms api=0ms
@933ms datastore_v3.Put real=257ms api=0ms

Notice how long it takes each PUT, this is not normal is it? The
objects I am updating are not overly large and I didn't have this
issue a few weeks ago - in fact I remember the same resource used to
have a typical request time of around 400ms not the 10,000-15,000ms it
currently takes.

Google employees, is this a known issue or is it a new one to add to
the recent problems?

On Sep 29, 10:35 pm, mscwd01 <mscw...@gmail.com> wrote:
> Hey,
>
> For the last week or so I have been getting a high error rate (approx
> 40%) on my Java app along with this error:
>
> Request was aborted after waiting too long to attempt to service your
> request because you have reached your simultaneous dynamic request
> limit. This is almost always due to excessively high latency in your

> app. Please seehttp://code.google.com/appengine/kb/general.html#activerequests

[Ikai Lan (Google)]

The latencies on those puts look a bit high, but not in the 10k ms range. Where are you seeing this number?

--

Ikai Lan 
Developer Programs Engineer, Google App Engine

Blogger: http://googleappengine.blogspot.com

Reddit: http://www.reddit.com/r/appengine

Twitter: http://twitter.com/app_engine

--
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.

[mscwd01]

The 10k ms is the overall latency of the request as quoted in my
dashboard logs. If you total each RPC the request normally takes
10-15k ms to complete.

GETs to the datastore are quick it's just the PUTs which are
significantly slowing the resource at present.

As a result I get hundreds of these in my logs:

#500 10350ms 0cpu_ms 0kb Test Harness,gzip(gfe)
[29/Sep/2010:16:13:35 -0700] "GET /ad/get HTTP/1.1" 500 0 - "Test
Harness,gzip(gfe)" "apvert.appspot.com" ms=10350 cpu_ms=0 api_cpu_ms=0
cpm_usd=0.000019 throttle_code=2

#W 09-29 04:13PM 35.067

Request was aborted after waiting too long to attempt to service your
request because you have reached your simultaneous dynamic request
limit. This is almost always due to excessively high latency in your

app. Please see http://code.google.com/appengine/kb/general.html#activerequests
for more details.

On Sep 30, 1:45 am, "Ikai Lan (Google)" <ikai.l+gro...@google.com>
wrote:

> > google-appengi...@googlegroups.com<google-appengine%2Bunsubscrib e...@googlegroups.com>
> > .