tomcat-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From bugzi...@apache.org
Subject DO NOT REPLY [Bug 40877] - JK1 and IIS6: Problem with multithreading / shared memory
Date Wed, 15 Nov 2006 17:32:01 GMT
DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG·
RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT
<http://issues.apache.org/bugzilla/show_bug.cgi?id=40877>.
ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND·
INSERTED IN THE BUG DATABASE.

http://issues.apache.org/bugzilla/show_bug.cgi?id=40877





------- Additional Comments From bjoern@syltonline.de  2006-11-15 09:32 -------
Thank you for the fix. I did a quick test on the AMD64 server and it looks 
really good. Multithread in an applicationpool and app pool recyling seems to 
work fine.
A few tests later, only a few small flaws appeared.

First, you get an (one!) error when die app-pool recycles. This is easy 
reproduceable. Recycle every 100 requests, you'll get one error like these 
every 100 requests. Sometimes 2 per cycle at the same time.

[Wed Nov 15 15:32:34 2006] [5536:4756] [error] 
HttpExtensionProc::jk_isapi_plugin.c (1102): could not get a worker for name P-
NG-TC3Hx
[Wed Nov 15 15:32:54 2006] [5568:6800] [error] 
HttpExtensionProc::jk_isapi_plugin.c (1102): could not get a worker for name P-
NG-TC3Hx

Also, rarely, a "WriteClient failed" appears. This didn't happen before on the 
JK2. But I'm not totally sure if this applies to this jk-version as origin.

[Wed Nov 15 17:16:48 2006] [0540:6520] [error] write::jk_isapi_plugin.c (656): 
WriteClient failed with 00002746
[Wed Nov 15 17:16:48 2006] [0540:6520] [info]  
ajp_process_callback::jk_ajp_common.c (1421): Writing to client aborted or 
client network problems
[Wed Nov 15 17:16:48 2006] [0540:6520] [info]  ajp_service::jk_ajp_common.c 
(1813): (P-NG-TC3H1) request failed, because of client write error without 
recovery in send loop attempt=0
[Wed Nov 15 17:16:48 2006] [0540:6520] [info]  service::jk_lb_worker.c (874): 
unrecoverable error 200, request failed. Client failed in the middle of 
request, we can't recover to another instance.

And still in the jkstatus, the config and stati apply only to this app pool or 
even web / virtual server. I guess this is by design, because you mentioned to 
change the shared mem to a per-vhost basis. But it is still not very handy. 
Say you want to drain a tomcat to do maintainance, on our farms you'd need to 
do it not only on 11 frontent webservers but also on 41 webs per server. 
That's not practicable. Hopefully in the future, there is a per-vhost an 
global section in jkstatus. But this would need globally initialized shared 
mem again :)

Also tested was the "reference" directive, which didn't work in the 1.2.19 on 
IIS. That also worked fine.

All in all, great work, thanks. If you can fix the last on-recycle error I'd 
say test passed for amd64. Any idea when the final 1.2.20 will be released?

-- 
Configure bugmail: http://issues.apache.org/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@tomcat.apache.org
For additional commands, e-mail: dev-help@tomcat.apache.org


Mime
View raw message