tomcat-users mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Rainer Jung <rainer.j...@kippdata.de>
Subject Re: Jakarta ISAPI errno=61
Date Tue, 14 Oct 2008 14:32:01 GMT
Matt Morten schrieb:
> We are currently experiencing an issue with Jakarta ISAPI Redirector 1.2.25,
> where our log is filling up with the following and our clients are receiving
> 503 errors:
> 
> [Thu Sep 25 12:05:28.937 2008] [732:3700] [info] jk_ajp_common.c (891):
> Failed opening socket to (127.0.0.1:8009) (errno=61)

61 = Winsock Error 10061 = Connection refused

> I know for a fact that Tomcat (our backend) is working, because you can
> connect without trouble to it directly (without Jakarta) and we get no
> issues. This leads me to believe that the Connector is running out of

Here you mean you can connect to the HTTP connector if the Tomcat? The
above indicates, you won't be able to connect to the AJP connector (port
8009).

> processes. Here is my workers.properties:
> 
> worker.wlb.type=lb
> worker.wlb.balance_workers=gauss41
> worker.gauss41.type=ajp13
> worker.gauss41.host=localhost
> worker.gauss41.port=8009
> worker.gauss41.connection_pool_size=300
> worker.gauss41.connection_pool_timeout=200
> worker.gauss41.lbfactor=1

I expect you configured more concurrency inside IIS (300) than you have
configured threads in the AJP connector of Tomcat? Which version of
Tomcat are you using, how does the Connector element in server.xml (AJP)
look like and which Java version are you using (also: 32Bit or 64Bit)?

Why do you think you actually need a maximum concurrency of 300?

> What is possibly more telling is the following messages we receive in normal
> operation in the ISAPI log file:
> 
> [Mon Oct 13 22:31:15.921 2008] [3188:7160] [info] jk_ajp_common.c (1511):
> Writing to client aborted or client network problems
> [Mon Oct 13 22:31:15.921 2008] [3188:7160] [info] jk_ajp_common.c (1996):
> (gauss41) request failed, because of client write error without recovery in
> send loop attempt=0
> [Mon Oct 13 22:31:15.921 2008] [3188:7160] [info] jk_lb_worker.c (1092):
> unrecoverable error 200, request failed. Client failed in the middle of
> request, we can't recover to another instance.
> [Mon Oct 13 22:31:15.921 2008] [3188:7160] [info] jk_isapi_plugin.c (1533):
> service() failed because client aborted connection
> [Mon Oct 13 22:31:15.937 2008] [3188:7360] [error] jk_isapi_plugin.c (754):
> WriteClient failed with 00002745
> 
> I know this suggests that the client has hit "Stop" or navigated away from
> the page, but we get *a lot* of these messages - sometimes 1 per second.

So how are your response times? If they are bad, that it really might be
users hitting stop or retry. Another possibility would be some
monitoring mechansim that doesn't wait for the actual response.

Regards,

Rainer

---------------------------------------------------------------------
To start a new topic, e-mail: users@tomcat.apache.org
To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
For additional commands, e-mail: users-help@tomcat.apache.org


Mime
View raw message