February 2019 - mod_cluster-issues

[JBoss JIRA] (MODCLUSTER-522) Memory leak in processing MCMP, wrong apr pool used for allocation

by Michal Karm Babacek (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-522?page=com.atlassian.jira.pl... ] Michal Karm Babacek commented on MODCLUSTER-522: ------------------------------------------------ Hello, [~chandrajboss9], currently we are distributing only Windows builds on https://ci.modcluster.io/job/mod_proxy_cluster-2.x-windows/ for Linux, you are going to have to compile mod_cluster yourself. It is fairly straightforward though: https://modcluster.io/examples/#apache-http-server-examples Cheers Karm > Memory leak in processing MCMP, wrong apr pool used for allocation > ------------------------------------------------------------------ > > Key: MODCLUSTER-522 > URL: https://issues.jboss.org/browse/MODCLUSTER-522 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final, 2.0.0.Alpha1 > Environment: Solaris 10 x86_64, RHEL 6, Fedora 24, httpd 2.4.6, httpd 2.4.20 > Reporter: Michal Karm Babacek > Assignee: Jean-Frederic Clere > Priority: Critical > Fix For: 1.3.5.Final > > Attachments: mod_cluster-mem.jpg, mod_cluster-mem1.jpg > > > There seems to be a wrong apr pool used for processing certain MCMP commands. We should use short life span pools for immediate processing and server-lifetime pools only for really persistent configuration. In the current state, with 20+ tomcat workers (1 alias and 1 context each) and virtually no client requests, we could see slow, but steady growth of heap allocated memory. > TODO: Investigate the offending logic, make sure we ain't using long-lived pools for immediate processing. > Originally discovered by: [~j_sykora] and [~jmsantuci] > Illustrative memory overview - with constant number of Tomcats: > -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 9 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-522) Memory leak in processing MCMP, wrong apr pool used for allocation

by CHANDRA SEKHARREDDY (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-522?page=com.atlassian.jira.pl... ] CHANDRA SEKHARREDDY commented on MODCLUSTER-522: ------------------------------------------------ Dear Lorenzo, we are facing an issue with wildfly-10.1.0.Final and mod_cluster-1.3.3-11.fc27.x86_64,httpd-2.4.6-45.el7.x86_64. when we add more than 2 nodes in domain controller mode, Memory is eaten up by httpd processes and causes termination of processes by RHEL 7.5 OS. Where can i get the mod_cluster package of this 1.3.5 final version.?? > Memory leak in processing MCMP, wrong apr pool used for allocation > ------------------------------------------------------------------ > > Key: MODCLUSTER-522 > URL: https://issues.jboss.org/browse/MODCLUSTER-522 > Project: mod_cluster > Issue Type: Bug > Components: Native (httpd modules) > Affects Versions: 1.3.1.Final, 2.0.0.Alpha1 > Environment: Solaris 10 x86_64, RHEL 6, Fedora 24, httpd 2.4.6, httpd 2.4.20 > Reporter: Michal Karm Babacek > Assignee: Jean-Frederic Clere > Priority: Critical > Fix For: 1.3.5.Final > > Attachments: mod_cluster-mem.jpg, mod_cluster-mem1.jpg > > > There seems to be a wrong apr pool used for processing certain MCMP commands. We should use short life span pools for immediate processing and server-lifetime pools only for really persistent configuration. In the current state, with 20+ tomcat workers (1 alias and 1 context each) and virtually no client requests, we could see slow, but steady growth of heap allocated memory. > TODO: Investigate the offending logic, make sure we ain't using long-lived pools for immediate processing. > Originally discovered by: [~j_sykora] and [~jmsantuci] > Illustrative memory overview - with constant number of Tomcats: > -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 9 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-673) Add support for multiple routes in JSESSIONID

by Jean-Frederic Clere (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-673?page=com.atlassian.jira.pl... ] Jean-Frederic Clere commented on MODCLUSTER-673: ------------------------------------------------ [~rhusar] to do that look in get_route_balancer() it look to first . (if ((route = strchr(sessionid, '.')) != NULL )) we need a clever logic there.... Note that the route is stored via apr_table_setn(r->subprocess_env, "BALANCER_SESSION_ROUTE", route); etc... we need a list... It don't remember how to handle a list in apr... the failover logic is find_best_byrequests()... find_best_byrequests() is going to be called for each tries, picking the next node/worker shouldn't be too difficult. See find_best_roundrobin() in modules/proxy/examples/mod_lbmethod_rr.c > Add support for multiple routes in JSESSIONID > --------------------------------------------- > > Key: MODCLUSTER-673 > URL: https://issues.jboss.org/browse/MODCLUSTER-673 > Project: mod_cluster > Issue Type: Enhancement > Components: Native (httpd modules) > Affects Versions: 1.4.0.Final > Reporter: Paul Ferraro > Assignee: Radoslav Husar > Priority: Major > > WildFly will soon be able to annotate the JSESSIONID using multiple routes. Thus the route should be parsed as a list of targets (instead of a single target). When the target route contains multiple entries, the reverse proxy should attempt to direct requests to the 1st active worker in the list. -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 10 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-318) Allow mod_cluster connector address to be configurable

by Radoslav Husar (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-318?page=com.atlassian.jira.pl... ] Radoslav Husar updated MODCLUSTER-318: -------------------------------------- Summary: Allow mod_cluster connector address to be configurable (was: Make connector address user configurable) > Allow mod_cluster connector address to be configurable > ------------------------------------------------------ > > Key: MODCLUSTER-318 > URL: https://issues.jboss.org/browse/MODCLUSTER-318 > Project: mod_cluster > Issue Type: Feature Request > Components: Core & Container Integration (Java) > Affects Versions: 1.2.1.Final > Reporter: Aleksandar Kostadinov > Assignee: Radoslav Husar > Priority: Major > > MODCLUSTER-168 introduced a change that broke some existing working configurations relying on using IP instead of a hostname. Also it is likely to be the cause of MODCLUSTER-91. > To have everybody happy, this needs to be made configurable, perhaps a system property that user can set to force a particular IP or hostname. It would also be very useful if there is a special value that makes the behavior same as before the changes in MODCLUSTER-168. For example setting the value to -1 would return the original behavior. -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 10 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-673) Add support for multiple routes in JSESSIONID

by Radoslav Husar (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-673?page=com.atlassian.jira.pl... ] Work on MODCLUSTER-673 started by Radoslav Husar. ------------------------------------------------- > Add support for multiple routes in JSESSIONID > --------------------------------------------- > > Key: MODCLUSTER-673 > URL: https://issues.jboss.org/browse/MODCLUSTER-673 > Project: mod_cluster > Issue Type: Enhancement > Components: Native (httpd modules) > Affects Versions: 1.4.0.Final > Reporter: Paul Ferraro > Assignee: Radoslav Husar > Priority: Major > > WildFly will soon be able to annotate the JSESSIONID using multiple routes. Thus the route should be parsed as a list of targets (instead of a single target). When the target route contains multiple entries, the reverse proxy should attempt to direct requests to the 1st active worker in the list. -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 10 months

1
0
0 / 0

[JBoss JIRA] (MODCLUSTER-530) Add a disabled-contexts feature for mod_cluster

by Radoslav Husar (Jira)

[ https://issues.jboss.org/browse/MODCLUSTER-530?page=com.atlassian.jira.pl... ] Radoslav Husar closed MODCLUSTER-530. ------------------------------------- Resolution: Rejected > Add a disabled-contexts feature for mod_cluster > ----------------------------------------------- > > Key: MODCLUSTER-530 > URL: https://issues.jboss.org/browse/MODCLUSTER-530 > Project: mod_cluster > Issue Type: Feature Request > Components: Core & Container Integration (Java) > Affects Versions: 1.3.3.Final > Reporter: Aaron Ogburn > Assignee: Radoslav Husar > Priority: Major > > Currently, if you want contexts disabled at start up, you can only have them all disabled with auto-enable-contexts="false". In some cases, you may want only specific apps disabled at start up. So it would be helpful to have a disabled-contexts parameter. This would accept a list of contexts similar to excluded-contexts, but the listed contexts in disabled-contexts would designate what contexts must be added to the balancer in a disabled state. -- This message was sent by Atlassian Jira (v7.12.1#712002)

5 years, 10 months

1
0
0 / 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

mod_cluster-issues February 2019