Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 9:41 PM, Razi wrote: -Original Message- From: Christopher Schultz Sent: Thursday, April 23, 2015 8:03 AM To: Tomcat Users List Subject: Re: Apache Tomcat jk connector 400 bad request Razi, On 4/22/15 6:39 PM, Razi Ansari wrote: Original message From: Christopher Schultz ch...@christopherschultz.net Date: 04/23/2015 6:15 AM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector. I tried inceasing the value of webserver KeepAliveTimeout to 15 and i dont get the 400 bad request error in the test environment. But i really want to know the root cause before i move it to production. In my httpwatch i see the request going without any body ony header. Without the change in KeepAliveTimeout, can you reproduce this error in your test environment? I would honestly be surprised if KeepAliveTimeout is the trouble, here. For the AJP connector, the KeepAliveTimeout is not relevant because AJP is by definition keep-alive. You can disable it if you think doing so will improve the situation, but I don't have any reason to suspect it would. It would be great to know whether the AJP connector or the HTTP connector was the one failing. In Tomcat, the stack trace will include the thread identifier which includes the port number and protocol being used. Please include that with your stack traces when you find the m. It would also be great to know which (exact) version of Tomcat is being used under the hood. Check the logs during startup to see what Tomcat says it is. -chris - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org Christopher, When KeepAliveTimeOut is 5, 12 , I get the error, when its 15 and above I don't get the error. That's ... weird. Are these particularly large requests? Are your clients on particularly slow connections? The versions of Apahce and Mod_jk are as follows Apache 2.2.24 Mod_jk 1.2.37 There is a more resent version of mod_jk, but that one should be okay. I'd upgrade if it's a possibility. I dont use Tomcat, my applicaton server is JBOSS EAP 6.1.0 I understand that. I can't remember when JBoss switched from using Tomcat as its servlet container to using their own in-house container, but older versions of JBoss use Tomcat internally, and that Tomcat has a version number. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOkQTAAoJEBzwKT+lPKRYY0IP/RdNIVopTKebt2G7L4S5xgh2 TQKDEhybN5pHuk4VI/xIttQaCVGIfH6tK0msLrD2GzcEqtQBDeXcjYuf6J3OStRW 5Xwm/3Pi2GPOcEiBYhbb0bhaqy7asmAIeuKXH4lwKJMl4JJQnIDy0Wp9dcLcODSv 2mTEE2jSLwdb8+j0DneB72oFNkAoKDEQkVqR+ajsFqicZCx4LBUAvkudGFgwuJzU nqDMCUKiq9hMso4BbiVYcVN4mf5eeE964JU6vf7Bt9mlA4ZZeiLnC52BdpeU2u3i MJHNHG6udDafupj10++wo3jvtWjBPtKiHbqFg0Fwl2cjm4fCAkEdj7VP80o7nrSY rApnFOt6l049TyWmYX0nmRFVRQxNHELkfpU8ZdvjYv1e0Q1OvZPu667GtD+zpFdz ar/bhcuni2CSOh9+zq5R03RPs2GdOVqiL1yHuHH4e+/diQKf7aJY0UJjczWoj2vS QYKRRIckDDHrA60iEs5XkaC0gM5/bwlCcgnlMppEcntAN64fPv47R+3FtRua0s3R Jo71pUiIK91cKG+sHXDd9DBSFVM82TU20BM7mdrK6kHdxo8h1cQrAdMwGxt0+iWR gJqoBBRJ5IxwBMoYlnVpWKdaiaZRssm/jU9wiwLYun4ZmuGOMwnoHzejMWmiTvt2 OkeF6mIuNrPSXqZKCXWD =ktm0 -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
hi Christopher, I checked up on the firewall, there is none between the webserver and the jboss application server. I had enabled the trace in modjk.log and found the following entries, with KeepAliveTimeout set to 5 Apr21 15:31:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common.c(1399): enter ... a bunch of other requests. ... Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common.c(1432): exit Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_into_msg_buff::jk_ajp_common.c(1487): exit Apr21 15:36:53 2015 [4023:140648883541760][debug]ajp_send_request::jk_ajp_common.c(1766): (worker12) browser stop sending data, no need to recover unrecoverable error 400 , request failed Consumed 0 bytes of remaining request data for worker aborting connection for worker attempting to map uri /error/http_bad_Request.html There is no error in the Jboss application server logs. Warm Regards Razi A. Ansari HP # 90625741 -Original Message- From: Razi Sent: Thursday, April 23, 2015 9:41 AM To: Tomcat Users List Subject: Re: Apache Tomcat jk connector 400 bad request Christopher, When KeepAliveTimeOut is 5, 12 , I get the error, when its 15 and above I don't get the error. The versions of Apahce and Mod_jk are as follows Apache 2.2.24 Mod_jk 1.2.37 I dont use Tomcat, my applicaton server is JBOSS EAP 6.1.0 I will get back to you with more logs. Thanks Warm Regards Razi A. Ansari -Original Message- From: Christopher Schultz Sent: Thursday, April 23, 2015 8:03 AM To: Tomcat Users List Subject: Re: Apache Tomcat jk connector 400 bad request -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 6:39 PM, Razi Ansari wrote: Original message From: Christopher Schultz ch...@christopherschultz.net Date: 04/23/2015 6:15 AM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector. I tried inceasing the value of webserver KeepAliveTimeout to 15 and i dont get the 400 bad request error in the test environment. But i really want to know the root cause before i move it to production. In my httpwatch i see the request going without any body ony header. Without the change in KeepAliveTimeout, can you reproduce this error in your test environment? I would honestly be surprised if KeepAliveTimeout is the trouble, here. For the AJP connector, the KeepAliveTimeout is not relevant because AJP is by definition keep-alive. You can disable it if you think doing so will improve the situation, but I don't have any reason to suspect it would. It would be great to know whether the AJP connector or the HTTP connector was the one failing. In Tomcat, the stack trace will include the thread identifier which includes the port number and protocol being used. Please include that with your stack traces when you find the m. It would also be great to know which (exact) version of Tomcat is being used under the hood. Check the logs during startup to see what Tomcat says it is. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVODbWAAoJEBzwKT+lPKRYiBkP+QENinLAv9vXdjDi598Av56n nin9n4jzkgVaMC7h/EpzM81w7rjVoNdfD9d3j472hQFyUY+x9iQwDxNRc25t7VHH jwq21AZBqRDn6hNvLxxpXyKqVuLImZ08GFzI4PdelIQ0IuZ2WlRQwW1xNtcVZKgh yq1/Az1zK82rzrIH8WJhu0frYNd5mqrrEEVCQfUs3pEJyqf0uQFM8buVVKScla3x sb6qjo3+XCHGZ/KXFllr0t0E7fGlk7xP8NhRLzSO7AkDvUgk7SzAryN2VmwbTn3w KbO3bzT7dQse+1ykj246L+TqIWJOsycooGagTRtH7kW/5jRXS6b8kLL08XhLY70z Ybcfbd1gwIbihm29LsOaU54hWTyQYa5cSAzCXEaQsGUfaoDsuU4L69Uf+G40QcO6 c1cCvJImSFjpmVcoVtVitT+U6vDDK0kIcuyrb766GCyde2pRlFswCMho2GN2yT6L 2e9v9xqmLANZE0iaJpK2qUVE/efh1m+API/wFYXqnQhe2S5V8fXAGXM/rJdeOQNL IXMZN4E1moVClbaKVn+TbtcexR9nmU6GPyaL+VoJauDJf5UV5Y31+wp+XgP1k13/ wQ0LoumpioPYp7QEbrSgNTUF5I2Q+Cp7VGtoO3PPDWzgcBH8eSiqHVnsC2da51k8 sjaBXvJn5/taUN49pEGo =0W9M -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, (What email program are you using? It doesn't seem to understand mailing lists because your replies don't include the thread-id required to properly-group mailing list threads. That's pretty frustrating because all your messages look separate from the others in the thread.) On 4/24/15 8:10 AM, Razi wrote: I checked up on the firewall, there is none between the webserver and the jboss application server. I had enabled the trace in modjk.log and found the following entries, with KeepAliveTimeout set to 5 What is KeepAliveTimeout? Is that your setting on httpd? If so, that only affects incoming requests from clients into httpd. It has no effect on the connections between httpd and Tomcat. Apr21 15:31:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common .c(1399): enter ... a bunch of other requests. ... Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common .c(1432): exit Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_into_msg_buff::jk_ajp_common.c(1 487): exit Apr21 15:36:53 2015 [4023:140648883541760][debug]ajp_send_request::jk_ajp_common.c(1766): (worker12) browser stop sending data, no need to recover unrecoverable error 400 , request failed Consumed 0 bytes of remaining request data for worker aborting connection for worker attempting to map uri /error/http_bad_Request.html There is no error in the Jboss application server logs. Interesting. What about the access log? Does JBoss even admit to accepting the request? - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOkHxAAoJEBzwKT+lPKRYrmAQAMBt4sgatAgrqHdSR9KxzYCo lUbdUjuQ+lGi/nCc/HALcDjVRTtbagSMsTcJxl5uzpUn9JbAwUN/+ERCh8mSMLyH jqIzrJD+lGJc+b5Eff6je3OFxZznpoieiN89oaodemVQzJW7oCSG5SNiYYqN6sBD cE01OE+V6sSohmqRNQ3Ieh4Maz4X9/J6qD2sq9ax+r5b71jZHeA6BK9//XOgolyV MeUSAuSC41ZnzdTMDozj8hrszilI5keBp4lWE0oIJJ7nHlkpqE0Zi13CN8NASUni Z2QQRu+SjVf5EHY9Cjek+qdr5z2HGdjQyxBIz7RW2a1EKH8jCU5f5tlb3Mjl2++k DywPlqfnJCeA03PPLuCnbiFW+M1M4JOoLrhKudwkKdsWz7SuURY1xy8G7ufue/ed 2d6hPINwVUb5Fr/FI7s2hj7ItZzx5IQ2DxSSCwdV3Ts0ePTK/acRKhu22lFfH0BH YnshsMkLrja7LlV5SpZeB9oCyPZl0ecoPnCy3K8sBdnDDGOddwflnSn97jgLRK+/ j1Rr9CT+DXQLgiWeFnJZJ5Fq5HCG1EzoidZ4KBeMWQpxGr1ukmhxDbMkWMUE87X2 ycG+GGAZ0x+xA0R8xyeN65hP/7cS+07FMHdKdw7rZAmf6BWyxC5FxFerVWRM2hPY IP4OKOfiXwivGBDwIgCC =ZBL9 -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
RE: Apache Tomcat jk connector 400 bad request
Date: Fri, 24 Apr 2015 09:15:29 -0400 From: ch...@christopherschultz.net To: users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, (What email program are you using? It doesn't seem to understand mailing lists because your replies don't include the thread-id required to properly-group mailing list threads. That's pretty frustrating because all your messages look separate from the others in the thread.) Apologies for the email, I am using Window Live Mail as my mail client. On 4/24/15 8:10 AM, Razi wrote: I checked up on the firewall, there is none between the webserver and the jboss application server. I had enabled the trace in modjk.log and found the following entries, with KeepAliveTimeout set to 5 What is KeepAliveTimeout? Is that your setting on httpd? If so, that only affects incoming requests from clients into httpd. It has no effect on the connections between httpd and Tomcat. KeepAliveTimeout, this is the one in httpd.conf of Apache Web Server. I set it to 5, i get the 400 error,increase it to 15, don't get the error. Apr21 15:31:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common .c(1399): enter ... a bunch of other requests. ... Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_fully_from_server::jk_ajp_common .c(1432): exit Apr21 15:36:53 2015 [4023:140648883541760][trace]ajp_read_into_msg_buff::jk_ajp_common.c(1 487): exit Apr21 15:36:53 2015 [4023:140648883541760][debug]ajp_send_request::jk_ajp_common.c(1766): (worker12) browser stop sending data, no need to recover unrecoverable error 400 , request failed Consumed 0 bytes of remaining request data for worker aborting connection for worker attempting to map uri /error/http_bad_Request.html There is no error in the Jboss application server logs. Interesting. What about the access log? Does JBoss even admit to accepting the request? The access log ,shows time of request received (%t), as 15:31:53 2015 , with 300 seconds as the time taken to serve the request (%D). But this line is actually printed 5 minutes afterward with other requests which were received at 15:36:48 , and also it shows 400 as status. Another thing I notice in the modjk.log, is that for this request, I only see the request header getting printed, don't see any body getting printed in the logs. On Jboss logs, I can see the request coming in, wait for 5 minutes and then process the request successfully. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOkHxAAoJEBzwKT+lPKRYrmAQAMBt4sgatAgrqHdSR9KxzYCo lUbdUjuQ+lGi/nCc/HALcDjVRTtbagSMsTcJxl5uzpUn9JbAwUN/+ERCh8mSMLyH jqIzrJD+lGJc+b5Eff6je3OFxZznpoieiN89oaodemVQzJW7oCSG5SNiYYqN6sBD cE01OE+V6sSohmqRNQ3Ieh4Maz4X9/J6qD2sq9ax+r5b71jZHeA6BK9//XOgolyV MeUSAuSC41ZnzdTMDozj8hrszilI5keBp4lWE0oIJJ7nHlkpqE0Zi13CN8NASUni Z2QQRu+SjVf5EHY9Cjek+qdr5z2HGdjQyxBIz7RW2a1EKH8jCU5f5tlb3Mjl2++k DywPlqfnJCeA03PPLuCnbiFW+M1M4JOoLrhKudwkKdsWz7SuURY1xy8G7ufue/ed 2d6hPINwVUb5Fr/FI7s2hj7ItZzx5IQ2DxSSCwdV3Ts0ePTK/acRKhu22lFfH0BH YnshsMkLrja7LlV5SpZeB9oCyPZl0ecoPnCy3K8sBdnDDGOddwflnSn97jgLRK+/ j1Rr9CT+DXQLgiWeFnJZJ5Fq5HCG1EzoidZ4KBeMWQpxGr1ukmhxDbMkWMUE87X2 ycG+GGAZ0x+xA0R8xyeN65hP/7cS+07FMHdKdw7rZAmf6BWyxC5FxFerVWRM2hPY IP4OKOfiXwivGBDwIgCC =ZBL9 -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/24/15 9:34 AM, Razi Ansari wrote: Date: Fri, 24 Apr 2015 09:15:29 -0400 From: ch...@christopherschultz.net To: users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request Razi, (What email program are you using? It doesn't seem to understand mailing lists because your replies don't include the thread-id required to properly-group mailing list threads. That's pretty frustrating because all your messages look separate from the others in the thread.) Apologies for the email, I am using Window Live Mail as my mail client. It's okay. It's just a minor irritation. But it /will/ make the archives a mess. On 4/24/15 8:10 AM, Razi wrote: I checked up on the firewall, there is none between the webserver and the jboss application server. I had enabled the trace in modjk.log and found the following entries, with KeepAliveTimeout set to 5 What is KeepAliveTimeout? Is that your setting on httpd? If so, that only affects incoming requests from clients into httpd. It has no effect on the connections between httpd and Tomcat. KeepAliveTimeout, this is the one in httpd.conf of Apache Web Server. I set it to 5, i get the 400 error,increase it to 15, don't get the error. Yeah... that's really weird. There is no error in the Jboss application server logs. Interesting. What about the access log? Does JBoss even admit to accepting the request? The access log ,shows time of request received (%t), as 15:31:53 2015 , with 300 seconds as the time taken to serve the request (%D). But this line is actually printed 5 minutes afterward with other requests which were received at 15:36:48 , and also it shows 400 as status. Another thing I notice in the modjk.log, is that for this request, I only see the request header getting printed, don't see any body getting printed in the logs. On Jboss logs, I can see the request coming in, wait for 5 minutes and then process the request successfully. The good news is that Tomcat is accepting the request, logging it, etc. A 400 response usually means that the request is broken in some way. It could be a partial request or something like that. For instance, an HTTP/1.1 request that never provides the \r\n\r\n required after the headers would just hang waiting or the \r\n\r\n. When the request-read timeout (keepAliveTimeout, defaulting to connectionTimeout, defaulting to -1; infinite) occurs, the connection will simply cancel the in-flight request. Have you set connectionTimeout or keepAliveTimeout to something other than their defaults? This would be in the JBoss configuration. I don't know how those timeouts are expressed in JBoss, but in Tomcat they would be on the Connector element. I think you might want to take this question to JBoss, especially if you are using a version of JBoss that doesn't use Tomcat under the cover s. Your mod_jk configuration looks fine to me; I can't think of a reason why you would be getting these dropped connections, unless you are under some kind of attack by someone trying to exploit a request-splitting vulnerability that exists somewhere in your stack. (And it would have to exist, because some component is convinced that there is a request that hasn't been fully made, and presumably the client is only sending complete requests.) You may have to pull-out a packet-sniffer for this one. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOk3CAAoJEBzwKT+lPKRYGSsQAJFkvk/lrhGT92Em2BRJfVnh uzfHA3h8+Ynbgcc3CJzLGYleXbvC/HSIzcC1YTPmZZGs8mOBTz3yicwRwbrPryC/ OlqC4v0lloN0eTZ8dub7P6dUg8g5awP+4G6r8WEdZ55WuIReQgDo3kD/2Md8+RXF rwPbOHFWUCwF56URqEEJ2fSbjH1D37lxT+oR6BCHGcRftoUmnffPCHXY5dY1RGNg k8tuvREPvPz6HE00JYpIfnphCnS7z37fo+fQgNmyXqwKhE8aWnQZEZ2R5zlK+u78 7ex04iSIGFJSrh3vOci6Vq9R5i3dIBKK2s/WGuUQ9aKcAbcxnyqgfkz9ssp6fWUi v4fQ+li1ZwwP9SYB6XgT6yGRwZ2UJsOGHfV93AkjzZ505vEJd6r1LjaB+ZxZ1Z6V P2Km9VrQe844QdNOGphWTYaEDYUXjLScSlS8gAXWG/zh9r0dUravZiiXReXY8FLd I6mkzK1+ThuOEQqvHDN+hN9ClrlSlQ9D6JbN90BdlhQQA/5PGlwyKluapG/4UEpe 0+vUALR6/IiQ4HwgrkitXGpOjPvVTZWnjs2pyS5ulezs5qzMNXdctKzWLsQ8tuR7 F/uosc575v1om/j2XKtrbK/+lNA2PSEJ9WhjLsL8dLygH1/jmL1LDJQwAeESzock gLQX0EYmGUG8QJBI1+sq =MlqE -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/21/15 5:18 PM, Razi wrote: Another bit of information I wanted to add. The apache error log is peppered with the following line : OpenSSL : I/O error, 5 bytes expected to read on BIO#... Are those errors correlated with the AJAX failures? Do non-AJAX requests fail in this way? Please copy/paste your Tomcat Connector configuration as well as your mod_jk properties for this worker. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVN7RlAAoJEBzwKT+lPKRY9swP/3zVEi9HE2Y5LshvuasAsk1R FiSty8XMgcsNcdM7DNG44TWHlxzaI8NjSxeoVbQAY1ENjGyz7PhkiRiD4CLBXV70 /DnlILOy/M5vT+O7pcdW0tHsF9XmeewkcN1VAef3fp3f+NbQU6U1xXpwDp+yWETy 7jOB0yBA/aqJPwkxqdfHd4BCi0+BvJBgUoSJ3vo67oYSA5lo0u/42zRWRgD1513k u14EJuYQfcevkwUMiX75KehvBqSj8O31zoYOBbeUjQSq2qshpJ+RKnDny1ZibHT5 ZfbtPEP9lvZChS/qdiz+vqHVQqPp67qnkDf4aO+TehSNyRfJ/vzMP93fwELWa4dB /AqAvOpxhN0bq9NN6vHLce5TY/5b6hhTGHuPqR8zXCS0VS8lgKGqkCJ5NmJfkAw/ Vdx0y60Uad01fCbSipT+/8X9zRpvOiaxvqix1s8NuiOHFdjDXEdnck0JdkWq4Wnm 2RfhZ9u9WixDEaS2lUQabSmHYxbB2gQBaL0kWQdttwaEKIaoXIry3IcEAreS8C9K ELloPXCR1qQXAv71Up31VYkK2jXmTE8QMIKrJOcd2hZPAQExKvVur2jTp/TFRa30 a2Qr2Gebnv9qwdJoM3bVTshgEROOMIMicOAGSn1juwOQ478u9TA2zKDJIALP7VyG 9/7QwAFFNxHMjghzzVXy =mWJG -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 6:39 PM, Razi Ansari wrote: Original message From: Christopher Schultz ch...@christopherschultz.net Date: 04/23/2015 6:15 AM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector. I tried inceasing the value of webserver KeepAliveTimeout to 15 and i dont get the 400 bad request error in the test environment. But i really want to know the root cause before i move it to production. In my httpwatch i see the request going without any body ony header. Without the change in KeepAliveTimeout, can you reproduce this error in your test environment? I would honestly be surprised if KeepAliveTimeout is the trouble, here. For the AJP connector, the KeepAliveTimeout is not relevant because AJP is by definition keep-alive. You can disable it if you think doing so will improve the situation, but I don't have any reason to suspect it would. It would be great to know whether the AJP connector or the HTTP connector was the one failing. In Tomcat, the stack trace will include the thread identifier which includes the port number and protocol being used. Please include that with your stack traces when you find the m. It would also be great to know which (exact) version of Tomcat is being used under the hood. Check the logs during startup to see what Tomcat says it is. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVODbWAAoJEBzwKT+lPKRYiBkP+QENinLAv9vXdjDi598Av56n nin9n4jzkgVaMC7h/EpzM81w7rjVoNdfD9d3j472hQFyUY+x9iQwDxNRc25t7VHH jwq21AZBqRDn6hNvLxxpXyKqVuLImZ08GFzI4PdelIQ0IuZ2WlRQwW1xNtcVZKgh yq1/Az1zK82rzrIH8WJhu0frYNd5mqrrEEVCQfUs3pEJyqf0uQFM8buVVKScla3x sb6qjo3+XCHGZ/KXFllr0t0E7fGlk7xP8NhRLzSO7AkDvUgk7SzAryN2VmwbTn3w KbO3bzT7dQse+1ykj246L+TqIWJOsycooGagTRtH7kW/5jRXS6b8kLL08XhLY70z Ybcfbd1gwIbihm29LsOaU54hWTyQYa5cSAzCXEaQsGUfaoDsuU4L69Uf+G40QcO6 c1cCvJImSFjpmVcoVtVitT+U6vDDK0kIcuyrb766GCyde2pRlFswCMho2GN2yT6L 2e9v9xqmLANZE0iaJpK2qUVE/efh1m+API/wFYXqnQhe2S5V8fXAGXM/rJdeOQNL IXMZN4E1moVClbaKVn+TbtcexR9nmU6GPyaL+VoJauDJf5UV5Y31+wp+XgP1k13/ wQ0LoumpioPYp7QEbrSgNTUF5I2Q+Cp7VGtoO3PPDWzgcBH8eSiqHVnsC2da51k8 sjaBXvJn5/taUN49pEGo =0W9M -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
Christopher, When KeepAliveTimeOut is 5, 12 , I get the error, when its 15 and above I don't get the error. The versions of Apahce and Mod_jk are as follows Apache 2.2.24 Mod_jk 1.2.37 I dont use Tomcat, my applicaton server is JBOSS EAP 6.1.0 I will get back to you with more logs. Thanks Warm Regards Razi A. Ansari -Original Message- From: Christopher Schultz Sent: Thursday, April 23, 2015 8:03 AM To: Tomcat Users List Subject: Re: Apache Tomcat jk connector 400 bad request -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 6:39 PM, Razi Ansari wrote: Original message From: Christopher Schultz ch...@christopherschultz.net Date: 04/23/2015 6:15 AM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector. I tried inceasing the value of webserver KeepAliveTimeout to 15 and i dont get the 400 bad request error in the test environment. But i really want to know the root cause before i move it to production. In my httpwatch i see the request going without any body ony header. Without the change in KeepAliveTimeout, can you reproduce this error in your test environment? I would honestly be surprised if KeepAliveTimeout is the trouble, here. For the AJP connector, the KeepAliveTimeout is not relevant because AJP is by definition keep-alive. You can disable it if you think doing so will improve the situation, but I don't have any reason to suspect it would. It would be great to know whether the AJP connector or the HTTP connector was the one failing. In Tomcat, the stack trace will include the thread identifier which includes the port number and protocol being used. Please include that with your stack traces when you find the m. It would also be great to know which (exact) version of Tomcat is being used under the hood. Check the logs during startup to see what Tomcat says it is. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVODbWAAoJEBzwKT+lPKRYiBkP+QENinLAv9vXdjDi598Av56n nin9n4jzkgVaMC7h/EpzM81w7rjVoNdfD9d3j472hQFyUY+x9iQwDxNRc25t7VHH jwq21AZBqRDn6hNvLxxpXyKqVuLImZ08GFzI4PdelIQ0IuZ2WlRQwW1xNtcVZKgh yq1/Az1zK82rzrIH8WJhu0frYNd5mqrrEEVCQfUs3pEJyqf0uQFM8buVVKScla3x sb6qjo3+XCHGZ/KXFllr0t0E7fGlk7xP8NhRLzSO7AkDvUgk7SzAryN2VmwbTn3w KbO3bzT7dQse+1ykj246L+TqIWJOsycooGagTRtH7kW/5jRXS6b8kLL08XhLY70z Ybcfbd1gwIbihm29LsOaU54hWTyQYa5cSAzCXEaQsGUfaoDsuU4L69Uf+G40QcO6 c1cCvJImSFjpmVcoVtVitT+U6vDDK0kIcuyrb766GCyde2pRlFswCMho2GN2yT6L 2e9v9xqmLANZE0iaJpK2qUVE/efh1m+API/wFYXqnQhe2S5V8fXAGXM/rJdeOQNL IXMZN4E1moVClbaKVn+TbtcexR9nmU6GPyaL+VoJauDJf5UV5Y31+wp+XgP1k13/ wQ0LoumpioPYp7QEbrSgNTUF5I2Q+Cp7VGtoO3PPDWzgcBH8eSiqHVnsC2da51k8 sjaBXvJn5/taUN49pEGo =0W9M -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
RE: Apache Tomcat jk connector 400 bad request
Please check inline for my reply . Thanks. Original message From: Christopher Schultz ch...@christopherschultz.net Date: 04/23/2015 6:15 AM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 5:45 PM, Razi wrote: Hi Christopher, Thanks for looking into this. Any random request fails, ajax or non-ajax. The worker.properties is as follows:: worker.lbroutex.type=lb worker.lbroutex.balance_workers=workerx,workery,workerz worker.lbroute.sticky_session=1 worker.workerX.port=1234 worker.workerX.host=$$$.com worker.workerX.type=ajp13 worker.workerX.lbfactor=1 As I am using jboss EAP, I am pasting the config from my standalone.xml subsystem xmlns=urn:jboss:domain:web:1.4 default-virtual-server=default-host instance-id=worker10 native=false connector name=http scheme=http protocol=HTTP/1.1 socket-binding=http/ connector name=https scheme=http protocol=HTTP/1.1 socket-binding=https secure=true ssl name=ssl key-alias=jboss password= certificate-key-file= protocol=/ /connector connector name=ajp scheme=http protocol=AJP/1.3 socket-binding=ajp/ I suppose this has a default port number (8009?) and it matches what you have in worker.workerX.port? Yes You shouldn't be using any OpenSSL for the the AJP connector, so proxied requests via AJP shouldn't trigger the OpenSSL errors. A full stack trace would be very helpful. I wiill get this trace out I don't see anything immediately obvious. Do you have a firewall between httpd and Tomcat? Has it been configured to leave the connections open forever? If not, you might want to consider configuring CPING/CPONG at intervals (look at the AJP connector configuration reference and search for cping/cpong to see how to do that ). will check for the firewall and revert. If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector.I tried inceasing the value of webserver KeepAliveTimeout to 15 and i dont get the 400 bad request error in the test environment. But i really want to know the root cause before i move it to production. In my httpwatch i see the request going without any body ony header. Any chance the 400 responses always come from the same httpd instance or Tomcat instance? No its random any instance. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOB1GAAoJEBzwKT+lPKRY51AQAJCuO+cuurN9CshykgXm+M2V P+oSLO3wCRrF8WhJtzDz/CfIGT6T679lQSAbMBNliVrxuu+Q+/UbAS4rHcAkO2Ou GRCb1xCX1qoSpl5qmppFsJMMMAZE7NhWP3ZkCC/FdRyj+Lb4ZJcuKmx8LaIEfuvF akWUhJD1sYAhclyYInpF78kXOyBcuP1/6dOtXtlUqZ0JuZiDGHEgdywVRJiZjzpz aTxSdz0AkL4/7svPBn5I6foV2vLUKindQjSn2L1Mjq5TsbEo3Dhe9xxSCL7dmK8u ZJ7Wp8Hi7Z3NrVMNf4YnLiChKhWN9mIhucMQMsY9nez7h2GO2P0LrVo5XRJV5R65 gzagTK1qSHjKrGJqvuBY757j3rsnG9jNEPVixg1IqIA/JIuC/CfO+eVsVsUJAQSH 3KuF3Ata098fBQAMxb9O4vegBm/JRjwJYeKXb+b+fjDIr6QFFgUBv8DNv7cultG9 zDVfZQ2vJqb+TBz5kf8gI5G0ZcwkL2+WveUt+GzvRRbAXiyTmCwIoMGDesi5VIx5 ojUcRWccXjKj6Cg3DKNqWw137/Cre+xVKgepkTjpPOQ0dyAK3G2cUuCgZhJy3OQj FPtfg4vXS5BJQLZCNVWktkLGaQqRwHTIeM7EAwvbaDh+290fZcg1vW99zqCGCLgt ksfiwrTNMLcEDPX//PrL =KsRp -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
Hi Christopher, Thanks for looking into this. Any random request fails, ajax or non-ajax. The worker.properties is as follows:: worker.lbroutex.type=lb worker.lbroutex.balance_workers=workerx,workery,workerz worker.lbroute.sticky_session=1 worker.workerX.port=1234 worker.workerX.host=$$$.com worker.workerX.type=ajp13 worker.workerX.lbfactor=1 As I am using jboss EAP, I am pasting the config from my standalone.xml subsystem xmlns=urn:jboss:domain:web:1.4 default-virtual-server=default-host instance-id=worker10 native=false connector name=http scheme=http protocol=HTTP/1.1 socket-binding=http/ connector name=https scheme=http protocol=HTTP/1.1 socket-binding=https secure=true ssl name=ssl key-alias=jboss password= certificate-key-file= protocol=/ /connector connector name=ajp scheme=http protocol=AJP/1.3 socket-binding=ajp/ virtual-server name=default-host enable-welcome-root=true alias name=localhost / alias name=example.com / /virtual-server /subsystem Warm Regards Razi A. Ansari -Original Message- From: Christopher Schultz Sent: Wednesday, April 22, 2015 10:47 PM To: Tomcat Users List Subject: Re: Apache Tomcat jk connector 400 bad request -BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/21/15 5:18 PM, Razi wrote: Another bit of information I wanted to add. The apache error log is peppered with the following line : OpenSSL : I/O error, 5 bytes expected to read on BIO#... Are those errors correlated with the AJAX failures? Do non-AJAX requests fail in this way? Please copy/paste your Tomcat Connector configuration as well as your mod_jk properties for this worker. - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVN7RlAAoJEBzwKT+lPKRY9swP/3zVEi9HE2Y5LshvuasAsk1R FiSty8XMgcsNcdM7DNG44TWHlxzaI8NjSxeoVbQAY1ENjGyz7PhkiRiD4CLBXV70 /DnlILOy/M5vT+O7pcdW0tHsF9XmeewkcN1VAef3fp3f+NbQU6U1xXpwDp+yWETy 7jOB0yBA/aqJPwkxqdfHd4BCi0+BvJBgUoSJ3vo67oYSA5lo0u/42zRWRgD1513k u14EJuYQfcevkwUMiX75KehvBqSj8O31zoYOBbeUjQSq2qshpJ+RKnDny1ZibHT5 ZfbtPEP9lvZChS/qdiz+vqHVQqPp67qnkDf4aO+TehSNyRfJ/vzMP93fwELWa4dB /AqAvOpxhN0bq9NN6vHLce5TY/5b6hhTGHuPqR8zXCS0VS8lgKGqkCJ5NmJfkAw/ Vdx0y60Uad01fCbSipT+/8X9zRpvOiaxvqix1s8NuiOHFdjDXEdnck0JdkWq4Wnm 2RfhZ9u9WixDEaS2lUQabSmHYxbB2gQBaL0kWQdttwaEKIaoXIry3IcEAreS8C9K ELloPXCR1qQXAv71Up31VYkK2jXmTE8QMIKrJOcd2hZPAQExKvVur2jTp/TFRa30 a2Qr2Gebnv9qwdJoM3bVTshgEROOMIMicOAGSn1juwOQ478u9TA2zKDJIALP7VyG 9/7QwAFFNxHMjghzzVXy =mWJG -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
-BEGIN PGP SIGNED MESSAGE- Hash: SHA256 Razi, On 4/22/15 5:45 PM, Razi wrote: Hi Christopher, Thanks for looking into this. Any random request fails, ajax or non-ajax. The worker.properties is as follows:: worker.lbroutex.type=lb worker.lbroutex.balance_workers=workerx,workery,workerz worker.lbroute.sticky_session=1 worker.workerX.port=1234 worker.workerX.host=$$$.com worker.workerX.type=ajp13 worker.workerX.lbfactor=1 As I am using jboss EAP, I am pasting the config from my standalone.xml subsystem xmlns=urn:jboss:domain:web:1.4 default-virtual-server=default-host instance-id=worker10 native=false connector name=http scheme=http protocol=HTTP/1.1 socket-binding=http/ connector name=https scheme=http protocol=HTTP/1.1 socket-binding=https secure=true ssl name=ssl key-alias=jboss password= certificate-key-file= protocol=/ /connector connector name=ajp scheme=http protocol=AJP/1.3 socket-binding=ajp/ I suppose this has a default port number (8009?) and it matches what you have in worker.workerX.port? You shouldn't be using any OpenSSL for the the AJP connector, so proxied requests via AJP shouldn't trigger the OpenSSL errors. A full stack trace would be very helpful. I don't see anything immediately obvious. Do you have a firewall between httpd and Tomcat? Has it been configured to leave the connections open forever? If not, you might want to consider configuring CPING/CPONG at intervals (look at the AJP connector configuration reference and search for cping/cpong to see how to do that ). If you decide to change your timeout values (I don't see a reason to do so unless you are encountering some kind of related problem, and I don't consider this to be a related problem), make sure they are consistent between mod_jk and Tomcat's AJP connector. Any chance the 400 responses always come from the same httpd instance or Tomcat instance? - -chris -BEGIN PGP SIGNATURE- Version: GnuPG v2 Comment: GPGTools - http://gpgtools.org iQIcBAEBCAAGBQJVOB1GAAoJEBzwKT+lPKRY51AQAJCuO+cuurN9CshykgXm+M2V P+oSLO3wCRrF8WhJtzDz/CfIGT6T679lQSAbMBNliVrxuu+Q+/UbAS4rHcAkO2Ou GRCb1xCX1qoSpl5qmppFsJMMMAZE7NhWP3ZkCC/FdRyj+Lb4ZJcuKmx8LaIEfuvF akWUhJD1sYAhclyYInpF78kXOyBcuP1/6dOtXtlUqZ0JuZiDGHEgdywVRJiZjzpz aTxSdz0AkL4/7svPBn5I6foV2vLUKindQjSn2L1Mjq5TsbEo3Dhe9xxSCL7dmK8u ZJ7Wp8Hi7Z3NrVMNf4YnLiChKhWN9mIhucMQMsY9nez7h2GO2P0LrVo5XRJV5R65 gzagTK1qSHjKrGJqvuBY757j3rsnG9jNEPVixg1IqIA/JIuC/CfO+eVsVsUJAQSH 3KuF3Ata098fBQAMxb9O4vegBm/JRjwJYeKXb+b+fjDIr6QFFgUBv8DNv7cultG9 zDVfZQ2vJqb+TBz5kf8gI5G0ZcwkL2+WveUt+GzvRRbAXiyTmCwIoMGDesi5VIx5 ojUcRWccXjKj6Cg3DKNqWw137/Cre+xVKgepkTjpPOQ0dyAK3G2cUuCgZhJy3OQj FPtfg4vXS5BJQLZCNVWktkLGaQqRwHTIeM7EAwvbaDh+290fZcg1vW99zqCGCLgt ksfiwrTNMLcEDPX//PrL =KsRp -END PGP SIGNATURE- - To unsubscribe, e-mail: users-unsubscr...@tomcat.apache.org For additional commands, e-mail: users-h...@tomcat.apache.org
Re: Apache Tomcat jk connector 400 bad request
On Wed, Apr 22, 2015 at 5:18 AM, Razi razi...@hotmail.com wrote: hi there, Another bit of information I wanted to add. The apache error log is peppered with the following line : OpenSSL : I/O error, 5 bytes expected to read on BIO#... Warm Regards Razi A. Ansari From: Razi Sent: Tuesday, April 21, 2015 8:37 AM To: users@tomcat.apache.org Subject: Apache Tomcat jk connector 400 bad request Hi there, I would like to explain my scenario, perhaps this has been answered on this forum. A bunch of random Ajax requests from the browser (IE9) end up with a 400 error code on the apache webserver and the the browser hangs for 5 minutes. Httpwatch shows the error code as ERROR_INTERNET_CONNECTION_RESET and then immediately afterwards IE fires the same request again, which shows up with a time taken of 5 minutes and error code as ERROR_HTTP_INVALID_SERVER_RESPONSE. The browser recovers after 5 minutes. Further investigation on the webserver and appserver logs reveals the following:: a.. The request comes from the browser and hits the webserver and then forwards to the appserver instantly. b.. The mod_jk log for the request shows that there is time duration of 5 minutes spent in the ajp_read_fully_server::jk_ajp_common.c(1399): enter. After 5 minutes I get the next line as follows ajp_read_fully_server::jk_ajp_common.c(1432): exit. Then in the next line i see the following ajp_send_request::jk_ajp_common.c(1766) worker 11 browser stop sending data, no need to recover. Later it shows unrecoverable 400, request failed. c.. The forensic.log show the content length as a nonzero value. d.. The applcation server log hangs in the org.apache.coyote.ajp.AjpProcessor.read method for 5 mintues and the continues the execution. The thread dump also confirms this. The questions I have are:: a.. Is this a problem with IE only because of the keepalive timeout and the apache webserver keepalive time(current value is set to 5seconds) out which is not in sync. b.. Is this a problem with the appserver not able to process requests that are bad/incomplete. c.. Should I increase the Apache webserver timeout value to 60s or more , will this have any performance impact. Kindly advise on the scenario. Many thanks for reading through. Current setup: Apache 2.2.24 Mod_jk 1.2.37 Redhat Linux VM JBoss EAP 6.1.0 JSF 2.1, Richfaces 3.3.4 First, I apologize if my comment is offtopic. But you may also consider fronting nginx instead of Apache httpd. It is very easy to configure as reverse proxy for your tomcat and also serve your static files. If Apache is really a requirement kindly ignore my comment Warm Regards Razi A. Ansari -- Marcos | I love PHP, Linux, and Java http://javadevnotes.com/java-string-length-examples
RE: Apache Tomcat jk connector 400 bad request
Hi Marcos Thanks for the comment. I am using apache to serve static content and it works well. Its just these random requests that fail with 400. I am trying out using a slightly higher value of the KeepAliveTimeout. I wanted to get some advise from the experts in this forum . Thanks. Original message From: Marcos Almeida Azevedo marcos.al.azev...@gmail.com Date: 04/22/2015 3:37 PM (GMT+08:00) To: Tomcat Users List users@tomcat.apache.org Subject: Re: Apache Tomcat jk connector 400 bad request On Wed, Apr 22, 2015 at 5:18 AM, Razi razi...@hotmail.com wrote: hi there, Another bit of information I wanted to add. The apache error log is peppered with the following line : OpenSSL : I/O error, 5 bytes expected to read on BIO#... Warm Regards Razi A. Ansari From: Razi Sent: Tuesday, April 21, 2015 8:37 AM To: users@tomcat.apache.org Subject: Apache Tomcat jk connector 400 bad request Hi there, I would like to explain my scenario, perhaps this has been answered on this forum. A bunch of random Ajax requests from the browser (IE9) end up with a 400 error code on the apache webserver and the the browser hangs for 5 minutes. Httpwatch shows the error code as ERROR_INTERNET_CONNECTION_RESET and then immediately afterwards IE fires the same request again, which shows up with a time taken of 5 minutes and error code as ERROR_HTTP_INVALID_SERVER_RESPONSE. The browser recovers after 5 minutes. Further investigation on the webserver and appserver logs reveals the following:: a.. The request comes from the browser and hits the webserver and then forwards to the appserver instantly. b.. The mod_jk log for the request shows that there is time duration of 5 minutes spent in the ajp_read_fully_server::jk_ajp_common.c(1399): enter. After 5 minutes I get the next line as follows ajp_read_fully_server::jk_ajp_common.c(1432): exit. Then in the next line i see the following ajp_send_request::jk_ajp_common.c(1766) worker 11 browser stop sending data, no need to recover. Later it shows unrecoverable 400, request failed. c.. The forensic.log show the content length as a nonzero value. d.. The applcation server log hangs in the org.apache.coyote.ajp.AjpProcessor.read method for 5 mintues and the continues the execution. The thread dump also confirms this. The questions I have are:: a.. Is this a problem with IE only because of the keepalive timeout and the apache webserver keepalive time(current value is set to 5seconds) out which is not in sync. b.. Is this a problem with the appserver not able to process requests that are bad/incomplete. c.. Should I increase the Apache webserver timeout value to 60s or more , will this have any performance impact. Kindly advise on the scenario. Many thanks for reading through. Current setup: Apache 2.2.24 Mod_jk 1.2.37 Redhat Linux VM JBoss EAP 6.1.0 JSF 2.1, Richfaces 3.3.4 First, I apologize if my comment is offtopic. But you may also consider fronting nginx instead of Apache httpd. It is very easy to configure as reverse proxy for your tomcat and also serve your static files. If Apache is really a requirement kindly ignore my comment Warm Regards Razi A. Ansari -- Marcos | I love PHP, Linux, and Java http://javadevnotes.com/java-string-length-examples
Re: Apache Tomcat jk connector 400 bad request
hi there, Another bit of information I wanted to add. The apache error log is peppered with the following line : OpenSSL : I/O error, 5 bytes expected to read on BIO#... Warm Regards Razi A. Ansari From: Razi Sent: Tuesday, April 21, 2015 8:37 AM To: users@tomcat.apache.org Subject: Apache Tomcat jk connector 400 bad request Hi there, I would like to explain my scenario, perhaps this has been answered on this forum. A bunch of random Ajax requests from the browser (IE9) end up with a 400 error code on the apache webserver and the the browser hangs for 5 minutes. Httpwatch shows the error code as ERROR_INTERNET_CONNECTION_RESET and then immediately afterwards IE fires the same request again, which shows up with a time taken of 5 minutes and error code as ERROR_HTTP_INVALID_SERVER_RESPONSE. The browser recovers after 5 minutes. Further investigation on the webserver and appserver logs reveals the following:: a.. The request comes from the browser and hits the webserver and then forwards to the appserver instantly. b.. The mod_jk log for the request shows that there is time duration of 5 minutes spent in the ajp_read_fully_server::jk_ajp_common.c(1399): enter. After 5 minutes I get the next line as follows ajp_read_fully_server::jk_ajp_common.c(1432): exit. Then in the next line i see the following ajp_send_request::jk_ajp_common.c(1766) worker 11 browser stop sending data, no need to recover. Later it shows unrecoverable 400, request failed. c.. The forensic.log show the content length as a nonzero value. d.. The applcation server log hangs in the org.apache.coyote.ajp.AjpProcessor.read method for 5 mintues and the continues the execution. The thread dump also confirms this. The questions I have are:: a.. Is this a problem with IE only because of the keepalive timeout and the apache webserver keepalive time(current value is set to 5seconds) out which is not in sync. b.. Is this a problem with the appserver not able to process requests that are bad/incomplete. c.. Should I increase the Apache webserver timeout value to 60s or more , will this have any performance impact. Kindly advise on the scenario. Many thanks for reading through. Current setup: Apache 2.2.24 Mod_jk 1.2.37 Redhat Linux VM JBoss EAP 6.1.0 JSF 2.1, Richfaces 3.3.4 Warm Regards Razi A. Ansari
Apache Tomcat jk connector 400 bad request
Hi there, I would like to explain my scenario, perhaps this has been answered on this forum. A bunch of random Ajax requests from the browser (IE9) end up with a 400 error code on the apache webserver and the the browser hangs for 5 minutes. Httpwatch shows the error code as ERROR_INTERNET_CONNECTION_RESET and then immediately afterwards IE fires the same request again, which shows up with a time taken of 5 minutes and error code as ERROR_HTTP_INVALID_SERVER_RESPONSE. The browser recovers after 5 minutes. Further investigation on the webserver and appserver logs reveals the following:: a.. The request comes from the browser and hits the webserver and then forwards to the appserver instantly. b.. The mod_jk log for the request shows that there is time duration of 5 minutes spent in the ajp_read_fully_server::jk_ajp_common.c(1399): enter. After 5 minutes I get the next line as follows ajp_read_fully_server::jk_ajp_common.c(1432): exit. Then in the next line i see the following ajp_send_request::jk_ajp_common.c(1766) worker 11 browser stop sending data, no need to recover. Later it shows unrecoverable 400, request failed. c.. The forensic.log show the content length as a nonzero value. d.. The applcation server log hangs in the org.apache.coyote.ajp.AjpProcessor.read method for 5 mintues and the continues the execution. The thread dump also confirms this. The questions I have are:: a.. Is this a problem with IE only because of the keepalive timeout and the apache webserver keepalive time(current value is set to 5seconds) out which is not in sync. b.. Is this a problem with the appserver not able to process requests that are bad/incomplete. c.. Should I increase the Apache webserver timeout value to 60s or more , will this have any performance impact. Kindly advise on the scenario. Many thanks for reading through. Current setup: Apache 2.2.24 Mod_jk 1.2.37 Redhat Linux VM JBoss EAP 6.1.0 JSF 2.1, Richfaces 3.3.4 Warm Regards Razi A. Ansari
