You are viewing a plain text version of this content. The canonical link for it is here.
Posted to users@tomcat.apache.org by Bocalinda <bo...@gmail.com> on 2009/09/30 13:08:47 UTC
Explanation mod_jk error messages
Hi List,
I have configured Apache 2.2.11 (worker mpm) with mod_jk 1.2.27 and two
nodes of Tomcat 6.0.20
My mod_jk configuration only has the following settings (apart from the port
and IP address settings ofcourse):
worker.tcnodo-11.socket_keepalive=true
worker.tcnodo-11.connection_pool_timeout=600
worker.tcnodo-12.socket_keepalive=true
worker.tcnodo-12.connection_pool_timeout=600
Inside the AJP connector tag of both Tomcats I have configure
connectionTimeout to 600000.
Now, my application works fine and seems to handle correctly failover of
failed nodes.
However, there are a series of errors and warnings printed in the mod_jk.log
which worry me.
I was wondering if somebody could explain what they actually mean and what
is happening.
I noticed that although connection failed to a node, it keeps trying sending
requests to that node over and over.
Although, my web application never returns an error (thus it seems its
balancing correctly to the working node), I'm worried about this behavior as
it might introduce latency.
[Wed Sep 30 11:31:21.339 2009] [4681:1183201600] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=2)
[Wed Sep 30 11:31:21.339 2009] [4681:1183201600] [error]
ajp_service::jk_ajp_common.c (2426): (tcnodo11-6eka) connecting to
tomcat failed.
[Wed Sep 30 11:31:21.339 2009] [4681:1183201600] [info]
service::jk_lb_worker.c (1347): service failed, worker tcnodo11-6eka is
in error state
[Wed Sep 30 11:31:57.244 2009] [4680:1357449536] [info]
ajp_send_request::jk_ajp_common.c (1456): (tcnodo12-6eka) all endpoints
are disconnected, detected by connect check (1), cping (0), send (0)
[Wed Sep 30 11:33:21.572 2009] [4769:1297041728] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:33:21.572 2009] [4769:1297041728] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:33:21.572 2009] [4769:1297041728] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:33:21.572 2009] [4769:1297041728] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=1)
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=2)
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [error]
ajp_service::jk_ajp_common.c (2426): (tcnodo11-6eka) connecting to
tomcat failed.
[Wed Sep 30 11:33:21.673 2009] [4769:1297041728] [info]
service::jk_lb_worker.c (1347): service failed, worker tcnodo11-6eka is
in error state
[Wed Sep 30 11:35:24.556 2009] [4769:1255082304] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:35:24.556 2009] [4769:1255082304] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:35:24.556 2009] [4769:1255082304] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:35:24.556 2009] [4769:1255082304] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=1)
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=2)
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [error]
ajp_service::jk_ajp_common.c (2426): (tcnodo11-6eka) connecting to
tomcat failed.
[Wed Sep 30 11:35:24.657 2009] [4769:1255082304] [info]
service::jk_lb_worker.c (1347): service failed, worker tcnodo11-6eka is
in error state
[Wed Sep 30 11:36:31.134 2009] [4769:1359980864] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:36:31.134 2009] [4769:1359980864] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:36:31.134 2009] [4769:1359980864] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:36:31.134 2009] [4769:1359980864] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=1)
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [info]
jk_open_socket::jk_connect.c (593): connect to 10.189.x.x:8009 failed
(errno=111)
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [info]
ajp_connect_to_endpoint::jk_ajp_common.c (922): Failed opening socket to
(10.189.x.x:8009) (errno=111)
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [error]
ajp_send_request::jk_ajp_common.c (1467): (tcnodo11-6eka) connecting to
backend failed. Tomcat is probably not started or is listening on the
wrong port (errno=111)
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [info]
ajp_service::jk_ajp_common.c (2407): (tcnodo11-6eka) sending request to
tomcat failed (recoverable), because of error during request sending
(attempt=2)
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [error]
ajp_service::jk_ajp_common.c (2426): (tcnodo11-6eka) connecting to
tomcat failed.
[Wed Sep 30 11:36:31.234 2009] [4769:1359980864] [info]
service::jk_lb_worker.c (1347): service failed, worker tcnodo11-6eka is
in error state
Re: Explanation mod_jk error messages
Posted by Bocalinda <bo...@gmail.com>.
Thanks a lot Rainer.
If the error messages are in some way "OK", then my question got answered.
Thanks again.
2009/9/30 Rainer Jung <ra...@kippdata.de>
> On 30.09.2009 17:19, Bocalinda wrote:
> > Hi André.
> >
> > I actually stopped one server on purpose, to test the mod_jk's behavior.
> > So the errors are not generated due to an overload or something.
> >
> > I would love to send you the whole workers.properties file and
> server.xml,
> > but unfurtunately at this moment I can't.
> > In advance I can tell you that both Tomncat's are running on diffferent
> > servers 10.189.x.x, on port 8009 and 8010 respectively.
> >
> > I stopped the server on port 8009, which coincides with the logs.
> Although,
> > it seems strange to me that it keeps trying to send request to that
> server.
> >
> > Hope this helps...
>
> Yes, if a backend node is gone and you don't inform mod_jk it will
> detect it, but it will also log it.
>
> Currently there is no separate probing of failed backends. If a backend
> has been detected as failed, mod_jk will use one of the requests which
> are eligible for this backend every now and then to see, if it works
> again. Depending on the load this should happen about once a minute.
> Usually mod_jk can detect very quickly, that the backend is still
> broken, log the fact and send the request to some other backend.
>
> If you to a backend out of service for some time, then it is nicer to
> set its attribute "activation" to stopped, Then mod_jk knows, that it
> shouldn't try to send anything there (and will not log any errors).
>
> Regards,
>
> Rainer
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: users-help@tomcat.apache.org
>
>
Re: Explanation mod_jk error messages
Posted by Rainer Jung <ra...@kippdata.de>.
On 30.09.2009 17:19, Bocalinda wrote:
> Hi André.
>
> I actually stopped one server on purpose, to test the mod_jk's behavior.
> So the errors are not generated due to an overload or something.
>
> I would love to send you the whole workers.properties file and server.xml,
> but unfurtunately at this moment I can't.
> In advance I can tell you that both Tomncat's are running on diffferent
> servers 10.189.x.x, on port 8009 and 8010 respectively.
>
> I stopped the server on port 8009, which coincides with the logs. Although,
> it seems strange to me that it keeps trying to send request to that server.
>
> Hope this helps...
Yes, if a backend node is gone and you don't inform mod_jk it will
detect it, but it will also log it.
Currently there is no separate probing of failed backends. If a backend
has been detected as failed, mod_jk will use one of the requests which
are eligible for this backend every now and then to see, if it works
again. Depending on the load this should happen about once a minute.
Usually mod_jk can detect very quickly, that the backend is still
broken, log the fact and send the request to some other backend.
If you to a backend out of service for some time, then it is nicer to
set its attribute "activation" to stopped, Then mod_jk knows, that it
shouldn't try to send anything there (and will not log any errors).
Regards,
Rainer
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
For additional commands, e-mail: users-help@tomcat.apache.org
Re: Explanation mod_jk error messages
Posted by Bocalinda <bo...@gmail.com>.
Hi André.
I actually stopped one server on purpose, to test the mod_jk's behavior.
So the errors are not generated due to an overload or something.
I would love to send you the whole workers.properties file and server.xml,
but unfurtunately at this moment I can't.
In advance I can tell you that both Tomncat's are running on diffferent
servers 10.189.x.x, on port 8009 and 8010 respectively.
I stopped the server on port 8009, which coincides with the logs. Although,
it seems strange to me that it keeps trying to send request to that server.
Hope this helps...
2009/9/30 André Warnier <aw...@ice-sa.com>
> Bocalinda wrote:
> ...
>
>
>> My mod_jk configuration only has the following settings (apart from the
>> port
>> and IP address settings ofcourse):
>>
>
> That is exactly what it would have been interesting to see, along with the
> <Connector> tags in your Tomcat's server.xml files.
>
> What is strange with the error messages you show, is that you never see an
> error in your applications. Maybe your back-end Tomcats are just overloaded
> from time to time, and have so many connection requests piled up that they
> start refusing new ones ?
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
> For additional commands, e-mail: users-help@tomcat.apache.org
>
>
Re: Explanation mod_jk error messages
Posted by André Warnier <aw...@ice-sa.com>.
Bocalinda wrote:
...
>
> My mod_jk configuration only has the following settings (apart from the port
> and IP address settings ofcourse):
That is exactly what it would have been interesting to see, along with
the <Connector> tags in your Tomcat's server.xml files.
What is strange with the error messages you show, is that you never see
an error in your applications. Maybe your back-end Tomcats are just
overloaded from time to time, and have so many connection requests piled
up that they start refusing new ones ?
---------------------------------------------------------------------
To unsubscribe, e-mail: users-unsubscribe@tomcat.apache.org
For additional commands, e-mail: users-help@tomcat.apache.org