November 2015 - mod_cluster-issues

[JBoss JIRA] (MODCLUSTER-353) SocketTimeoutException in DefaultMCMPHandler when using httpd 2.4.x

by Markus Locher (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-353?page=com.atlassian.jira.pl... ] Markus Locher updated MODCLUSTER-353: ------------------------------------- Attachment: screenshot-1.png > SocketTimeoutException in DefaultMCMPHandler when using httpd 2.4.x > ------------------------------------------------------------------- > > Key: MODCLUSTER-353 > URL: https://issues.jboss.org/browse/MODCLUSTER-353 > Project: mod_cluster > Issue Type: Bug > Affects Versions: 1.2.1.Final > Environment: httpd-2.4.6 and mod_cluster-1.2.1.Final and above > Reporter: Marco Danti > Assignee: Jean-Frederic Clere > Attachments: screenshot-1.png > > > After upgrading a web server from Apache 2.2.x (with mod_cluster 1.2.0) to Apache 2.4.6 (with mod_cluster 1.2.1.Final), when I bring up JBoss (7.2.0.Final) I start seeing repeated errors like the following: > > 14:45:36,777 INFO [org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler] (ContainerBackgroundProcessor[StandardEngine[jboss.web]]) IO error sending command CONFIG to proxy stw-1.priv.softeco.it/172.16.0.133:6666: java.net.SocketTimeoutException: Read timed out > at java.net.SocketInputStream.socketRead0(Native Method) [rt.jar:1.7.0_21] > at java.net.SocketInputStream.read(SocketInputStream.java:150) [rt.jar:1.7.0_21] > at java.net.SocketInputStream.read(SocketInputStream.java:121) [rt.jar:1.7.0_21] > at sun.nio.cs.StreamDecoder.readBytes(StreamDecoder.java:283) [rt.jar:1.7.0_21] > at sun.nio.cs.StreamDecoder.implRead(StreamDecoder.java:325) [rt.jar:1.7.0_21] > at sun.nio.cs.StreamDecoder.read(StreamDecoder.java:177) [rt.jar:1.7.0_21] > at java.io.InputStreamReader.read(InputStreamReader.java:184) [rt.jar:1.7.0_21] > at java.io.BufferedReader.fill(BufferedReader.java:154) [rt.jar:1.7.0_21] > at java.io.BufferedReader.readLine(BufferedReader.java:317) [rt.jar:1.7.0_21] > at java.io.BufferedReader.readLine(BufferedReader.java:382) [rt.jar:1.7.0_21] > at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequest(DefaultMCMPHandler.java:670) > at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.sendRequests(DefaultMCMPHandler.java:437) > at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.status(DefaultMCMPHandler.java:385) > at org.jboss.modcluster.mcmp.impl.DefaultMCMPHandler.status(DefaultMCMPHandler.java:349) > at org.jboss.modcluster.ModClusterService.status(ModClusterService.java:468) > at org.jboss.modcluster.container.catalina.CatalinaEventHandlerAdapter.lifecycleEvent(CatalinaEventHandlerAdapter.java:244) > at org.apache.catalina.util.LifecycleSupport.fireLifecycleEvent(LifecycleSupport.java:115) [jbossweb-7.2.0.Final.jar:7.2.0.Final] > at org.apache.catalina.core.ContainerBase.backgroundProcess(ContainerBase.java:1323) [jbossweb-7.2.0.Final.jar:7.2.0.Final] > at org.apache.catalina.core.ContainerBase$ContainerBackgroundProcessor.processChildren(ContainerBase.java:1588) [jbossweb-7.2.0.Final.jar:7.2.0.Final] > at org.apache.catalina.core.ContainerBase$ContainerBackgroundProcessor.run(ContainerBase.java:1574) [jbossweb-7.2.0.Final.jar:7.2.0.Final] > at java.lang.Thread.run(Thread.java:722) [rt.jar:1.7.0_21] -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-471) Mod_cluster performance on httpd 2.4

by Coty Sutherland (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-471?page=com.atlassian.jira.pl... ] Coty Sutherland commented on MODCLUSTER-471: -------------------------------------------- Yet another thing that might help, the user was running 2 CPUs for httpd with httpd 2.4 and claimed that the same configuration/hardware worked fine for httpd 2.2. They even increased the CPU count to 8 (VM limit) and the requests were still slow. I think that they were DoSing httpd with all those nodes with the default 10 second STATUS interval, but even after increasing that to 60 seconds, the problem persisted. > Mod_cluster performance on httpd 2.4 > ------------------------------------ > > Key: MODCLUSTER-471 > URL: https://issues.jboss.org/browse/MODCLUSTER-471 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final > Reporter: Coty Sutherland > Assignee: Jean-Frederic Clere > Attachments: reproducer.tar.gz > > > When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. > One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. I'm attaching the reproduction script that the user provided. It needs a receiver (a mod_cluster node running on localhost port 8009) and an httpd balancer. Then you can run the create script to create the nodes and the update script to update them using netcat to prevent the need for 250 real instances. It is a bit hacky, but it could save you some resources. -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-471) Mod_cluster performance on httpd 2.4

by Coty Sutherland (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-471?page=com.atlassian.jira.pl... ] Coty Sutherland updated MODCLUSTER-471: --------------------------------------- Attachment: reproducer.tar.gz Hacky reproducer from the user. > Mod_cluster performance on httpd 2.4 > ------------------------------------ > > Key: MODCLUSTER-471 > URL: https://issues.jboss.org/browse/MODCLUSTER-471 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final > Reporter: Coty Sutherland > Assignee: Jean-Frederic Clere > Attachments: reproducer.tar.gz > > > When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. > One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. I'm attaching the reproduction script that the user provided. It needs a receiver (a mod_cluster node running on localhost port 8009) and an httpd balancer. Then you can run the create script to create the nodes and the update script to update them using netcat to prevent the need for 250 real instances. It is a bit hacky, but it could save you some resources. -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-471) Mod_cluster performance on httpd 2.4

by Coty Sutherland (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-471?page=com.atlassian.jira.pl... ] Coty Sutherland updated MODCLUSTER-471: --------------------------------------- Description: When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. I'm attaching the reproduction script that the user provided. It needs a receiver (a mod_cluster node running on localhost port 8009) and an httpd balancer. Then you can run the create script to create the nodes and the update script to update them using netcat to prevent the need for 250 real instances. It is a bit hacky, but it could save you some resources. was: When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. > Mod_cluster performance on httpd 2.4 > ------------------------------------ > > Key: MODCLUSTER-471 > URL: https://issues.jboss.org/browse/MODCLUSTER-471 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final > Reporter: Coty Sutherland > Assignee: Jean-Frederic Clere > Attachments: reproducer.tar.gz > > > When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. > One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. I'm attaching the reproduction script that the user provided. It needs a receiver (a mod_cluster node running on localhost port 8009) and an httpd balancer. Then you can run the create script to create the nodes and the update script to update them using netcat to prevent the need for 250 real instances. It is a bit hacky, but it could save you some resources. -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-471) Mod_cluster performance on httpd 2.4

by Coty Sutherland (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-471?page=com.atlassian.jira.pl... ] Coty Sutherland updated MODCLUSTER-471: --------------------------------------- Description: When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. was:When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. > Mod_cluster performance on httpd 2.4 > ------------------------------------ > > Key: MODCLUSTER-471 > URL: https://issues.jboss.org/browse/MODCLUSTER-471 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final > Reporter: Coty Sutherland > Assignee: Jean-Frederic Clere > > When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. > One more important note is that the user is fronting about 250+ (up to 290) JBoss instances behind a single balancer. -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-471) Mod_cluster performance on httpd 2.4

by Coty Sutherland (JIRA)

Coty Sutherland created MODCLUSTER-471: ------------------------------------------ Summary: Mod_cluster performance on httpd 2.4 Key: MODCLUSTER-471 URL: https://issues.jboss.org/browse/MODCLUSTER-471 Project: mod_cluster Issue Type: Bug Components: Native (httpd modules) Affects Versions: 1.3.1.Final Reporter: Coty Sutherland Assignee: Jean-Frederic Clere When using mod_cluster with httpd 2.4 (specifically with the JWS 3.0.1 product family) in front of EAP 6.3 there is a noticeable performance decrease in requests being serviced from httpd. This seemingly is resolved by using the latest (at presently 1.3.2.Alpha), but research hasn't pointed to a specific fix there, so I'm opening this. We were thinking that it had something to do with module loading order, but haven't been able to confirm that. -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-470) Documentation states stickySessionForce defaults to true in modcluster subsystem and defaults to false in Proxy Configuration

by Robert Bost (JIRA)

[ https://issues.jboss.org/browse/MODCLUSTER-470?page=com.atlassian.jira.pl... ] Robert Bost updated MODCLUSTER-470: ----------------------------------- Component/s: Documentation & Demos Description: http://docs.jboss.org/mod_cluster/1.3.0/html_single/ shows conflicting explanations of stickySessionForce defaults Summary: Documentation states stickySessionForce defaults to true in modcluster subsystem and defaults to false in Proxy Configuration (was: Documentation states sticky) Affects Version/s: 1.3.0.Final Affects: Documentation (Ref Guide, User Guide, etc.) > Documentation states stickySessionForce defaults to true in modcluster subsystem and defaults to false in Proxy Configuration > ----------------------------------------------------------------------------------------------------------------------------- > > Key: MODCLUSTER-470 > URL: https://issues.jboss.org/browse/MODCLUSTER-470 > Project: mod_cluster > Issue Type: Bug > Components: Documentation & Demos > Affects Versions: 1.3.0.Final > Reporter: Robert Bost > Assignee: Jean-Frederic Clere > > http://docs.jboss.org/mod_cluster/1.3.0/html_single/ shows conflicting explanations of stickySessionForce defaults -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-470) Documentation states sticky

by Robert Bost (JIRA)

Robert Bost created MODCLUSTER-470: -------------------------------------- Summary: Documentation states sticky Key: MODCLUSTER-470 URL: https://issues.jboss.org/browse/MODCLUSTER-470 Project: mod_cluster Issue Type: Bug Reporter: Robert Bost Assignee: Jean-Frederic Clere -- This message was sent by Atlassian JIRA (v6.4.11#64026)

8 years, 5 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

mod_cluster-issues November 2015