I get bursty periods of very high request latency, with increasing
frequency over the last few weeks, which last from seconds to hours.
Usually the elevated latencies are in the 3-6s range but sometimes it
goes up into the 20-30s range. For example, here are some logs from a
few minutes ago:
09-18 12:23AM 46.511 /checkStatus 500 2906ms 105cpu_ms
8api_cpu_ms 0kb Pingdom.com_bot_version_1.4_(http://
www.pingdom.com/),gzip(gfe),gzip(gfe)
09-18 12:23AM 46.511 /checkStatus 500 2906ms 105cpu_ms
8api_cpu_ms 0kb Pingdom.com_bot_version_1.4_(http://
www.pingdom.com/),gzip(gfe),gzip(gfe)
09-18 12:23AM 46.360 /stats 200 2771ms 105cpu_ms 8api_cpu_ms 5kb
w3m/0.5.2,gzip(gfe),gzip(gfe)
09-18 12:23AM 46.118 /admin/tick 200 2706ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
09-18 12:23AM 43.878 /admin/tick 200 3825ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
09-18 12:23AM 41.639 /admin/tick 200 4983ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
09-18 12:23AM 41.603 /stats 200 4602ms 105cpu_ms 8api_cpu_ms 5kb
w3m/0.5.2,gzip(gfe),gzip(gfe)
09-18 12:23AM 41.603 /stats 200 4602ms 105cpu_ms 8api_cpu_ms 5kb
w3m/0.5.2,gzip(gfe),gzip(gfe)
09-18 12:23AM 39.401 /admin/tick 200 4826ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
09-18 12:23AM 39.401 /admin/tick 200 4826ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
09-18 12:23AM 37.158 /admin/tick 200 4800ms 113cpu_ms
16api_cpu_ms 0kb Python-urllib/2.5,gzip(gfe),gzip(gfe)
Pingdom is pinging from all over the globe and sees the problem so it
is definitely on the App Engine side.
Drilling down into my logs: Even when the requests are taking 6s+ to
return, from my first logging statement at the very top of my python
file to the last logging statement at the end of my request handler is
only 500ms of real time. So, I don't know where exactly the issue is,
but it doesn't seem to be inside my code.
These requests all have < 200ms of latency most of the time (although
the bad behavior has gotten increasingly frequent lately).
I thought this might be related to the problems the Django folks are
having with DeadlineExceeded errors... See
http://code.google.com/p/googleappengine/issues/detail?id=1695 and
many other discussion threads linked therein.
--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the Google Groups
"Google App Engine" group.
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to
[email protected]
For more options, visit this group at
http://groups.google.com/group/google-appengine?hl=en
-~----------~----~----~----~------~----~------~--~---
