It seems to me that the time harakiri should be counted from the moment of
receipt of the request, and not from the start of its execution.
If I set the harakiri at 3, I hope that no request will be made longer than
3s. And if they arise, they will see it in the logs.
Yet this is not true. eg.
-p 1
$ ab -c 8 -n 10 localhost
...
100% 16025 (longest request)
So there are requests 16s at harakiri 3 and after this there is no trace in the
logs.
What's more ...
[pid: 8542|app: 0|req: 5/5] ::ffff:127.0.0.1 () {30 vars in 379 bytes} [Sun Feb
5 18:01:01 2012] GET / => generated 2 bytes in 2002 msecs (HTTP/1.0 200) 2
headers in 78 bytes (1 switches on core 0)
[pid: 8542|app: 0|req: 6/6] ::ffff:127.0.0.1 () {30 vars in 379 bytes} [Sun Feb
5 18:01:03 2012] GET / => generated 2 bytes in 2002 msecs (HTTP/1.0 200) 2
headers in 78 bytes (1 switches on core 0)
[pid: 8542|app: 0|req: 7/7] ::ffff:127.0.0.1 () {30 vars in 379 bytes} [Sun Feb
5 18:01:05 2012] GET / => generated 2 bytes in 2003 msecs (HTTP/1.0 200) 2
headers in 78 bytes (1 switches on core 0)
Logs suggest that each of these tasks was to wait 2 seconds, which is not true.
sneaky, sneaky ;)
--
Łukasz Wróblewski
http://www.nri.pl/ - Nowoczesne Rozwiązania Internetowe
http://www.hostowisko.pl/ - Profesjonalny i tani hosting
http://www.katalog-polskich-firm.pl/ - Najlepszy darmowy katalog firm
_______________________________________________
uWSGI mailing list
[email protected]
http://lists.unbit.it/cgi-bin/mailman/listinfo/uwsgi
