Server: Centos 6.4 64 bit / Spacewalk 1.9 VMWare When syncing large repositories (Fedora 18/Centos 6 base), I always get this message in the repo log: [Errno 32] Broken pipe ERROR: [Errno 32] Broken pipe Sometimes taskomatic also stops working and needs a restart/start. I am starting the sync via the web interface. Additional info: I am running Spacewalk on a VM (3GB Memory, PLSQL, slow HDD) Taskometer Log INFO | jvm 4 | 2013/05/21 08:26:22 | May 21, 2013 8:26:21 AM com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector run INFO | jvm 4 | 2013/05/21 08:26:22 | WARNING: com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector@60f00e0f -- APPARENT DEADLOCK!!! Creating emergency threads for unassigned pending tasks! INFO | jvm 4 | 2013/05/21 08:26:32 | May 21, 2013 8:26:32 AM com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector run INFO | jvm 4 | 2013/05/21 08:26:32 | WARNING: com.mchange.v2.async.ThreadPoolAsynchronousRunner$DeadlockDetector@60f00e0f -- APPARENT DEADLOCK!!! Complete Status: INFO | jvm 4 | 2013/05/21 08:26:32 | Managed Threads: 3 INFO | jvm 4 | 2013/05/21 08:26:32 | Active Threads: 0 INFO | jvm 4 | 2013/05/21 08:26:32 | Active Tasks: INFO | jvm 4 | 2013/05/21 08:26:32 | Pending Tasks: INFO | jvm 4 | 2013/05/21 08:26:32 | com.mchange.v2.resourcepool.BasicResourcePool$1RefurbishCheckinResourceTask@1d05c9a1 INFO | jvm 4 | 2013/05/21 08:26:32 | com.mchange.v2.resourcepool.BasicResourcePool$1RefurbishCheckinResourceTask@66f877e9 INFO | jvm 4 | 2013/05/21 08:26:32 | Pool thread stack traces: INFO | jvm 4 | 2013/05/21 08:26:32 | Thread[com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#1,5,main] INFO | jvm 4 | 2013/05/21 08:26:32 | java.lang.Object.wait(Native Method) INFO | jvm 4 | 2013/05/21 08:26:32 | com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread.run(ThreadPoolAsynchronousRunner.java:534) INFO | jvm 4 | 2013/05/21 08:26:32 | Thread[com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#0,5,main] INFO | jvm 4 | 2013/05/21 08:26:32 | java.lang.Object.wait(Native Method) INFO | jvm 4 | 2013/05/21 08:26:32 | com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread.run(ThreadPoolAsynchronousRunner.java:534) INFO | jvm 4 | 2013/05/21 08:26:32 | Thread[com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread-#2,5,main] INFO | jvm 4 | 2013/05/21 08:26:32 | java.lang.Object.wait(Native Method) INFO | jvm 4 | 2013/05/21 08:26:32 | com.mchange.v2.async.ThreadPoolAsynchronousRunner$PoolThread.run(ThreadPoolAsynchronousRunner.java:534) INFO | jvm 4 | 2013/05/21 08:26:32 | INFO | jvm 4 | 2013/05/21 08:26:32 |
Are there any errors in /var/log/rhn/reposync/*.log?
Hi Michael, the error was just the one mentioned above: [Errno 32] Broken pipe ERROR: [Errno 32] Broken pipe Here the last few lines of a few: [root@control reposync]# cat fedora18-x86_64-2013.05.20-17\:21\:40.log Sync started: Mon May 20 17:21:40 2013 ['/usr/bin/spacewalk-repo-sync', '--channel', 'fedora18-x86_64', '--type', 'yum'] Repo URL: https://mirrors.fedoraproject.org/metalink?repo=fedora-18&arch=x86_64 Packages in repo: 33868 Packages already synced: 0 ERROR: [Errno 32] Broken pipe [root@control reposync]# tail centos6-x86_64-2013.05.21-13\:54\:58.log 1114/6381 : python-ipaddr-2.1.9-3.el6-0.noarch 1115/6381 : libreadline-java-javadoc-0.8.0-24.3.el6-0.x86_64 1116/6381 : bison-2.4.1-5.el6-0.x86_64 1117/6381 : libevent-devel-1.4.13-4.el6-0.i686 1118/6381 : atmel-firmware-1.3-7.el6-0.noarch 1119/6381 : mythes-ru-0.20070613-3.1.el6-0.noarch 1120/6381 : vigra-devel-1.6.0-2.1.el6-0.i686 1121/6381 : meanwhile-devel-1.1.0-3.el6-0.x86_64 [Errno 32] Broken pipe ERROR: [Errno 32] Broken pipe [root@control reposync]# tail fedora18-i386-updates-2013.05.16-23\:33\:36.log failure: texlive-cmbright-svn21107.8.1-20.fc18.noarch.rpm from fedora18-i386-updates: [Errno 256] No more mirrors to try. 3781/12846 : uim-skk-1.8.5-2.fc18-0.i686 failure: uim-skk-1.8.5-2.fc18.i686.rpm from fedora18-i386-updates: [Errno 256] No more mirrors to try. 3782/12846 : gvfs-smb-1.14.2-3.fc18-0.i686 failure: gvfs-smb-1.14.2-3.fc18.i686.rpm from fedora18-i386-updates: [Errno 256] No more mirrors to try. 3783/12846 : libreoffice-ogltrans-3.6.6.2-5.fc18-1.i686 failure: libreoffice-ogltrans-3.6.6.2-5.fc18.i686.rpm from fedora18-i386-updates: [Errno 256] No more mirrors to try. 3784/12846 : texlive-uptex-bin-svn26912.0-20.20130321_r29448.fc18-2.i686 failure: texlive-uptex-bin-svn26912.0-20.20130321_r29448.fc18.i686.rpm from fedora18-i386-updates: [Errno 256] No more mirrors to try. ERROR: [Errno 32] Broken pipe
Could this have something to do with multiple syncs happening at the same time? I noticed from the logs, that when I restart Taskomatic, non-executed scheduled syncs also start running at some point. I have disabled the scheduled job and started the crashed Taskomatic again and now the download seems to be running for a lot longer (at least it managed over night).
This happens when the HDD is a network drive. I have moved the HDD to the local machine and it works fine. So the error probably shows up when the HDD transfer is too slow
This seem to be caused by incorrect handling of stdout and stderr of spacewalk-repo-sync. I'm closing it as duplicate of bug 820612. If you disagree please reopen the bug. *** This bug has been marked as a duplicate of bug 820612 ***
This BZ closed some time during 2.5, 2.6 or 2.7. Adding to 2.7 tracking bug.