Unison eating up system resources

Discussion in 'Installation/Configuration' started by tek, Jun 8, 2012.

  1. tek

    tek New Member

    I followed the how-TO for "Installing A Web, Email & MySQL Database Cluster On Debian 6.0 With ISPConfig 3" and have had it running for a while now one two blade servers. Everything was good for a while but now unison just goes crazy and no clue really why or what I can do about it.

    This is my last run on top with one of the blades just having been rebooted but in a few hours it will run out of memory, and start using swap and after that at some point services will start failing till I kill off unison and then it repeats.
    root@blade007:/etc# top
    top - 19:39:27 up 1:27, 2 users, load average: 8.69, 8.77, 7.84
    Tasks: 199 total, 11 running, 188 sleeping, 0 stopped, 0 zombie
    Cpu(s): 98.2%us, 0.9%sy, 0.0%ni, 0.0%id, 0.0%wa, 0.0%hi, 0.9%si, 0.0%st
    Mem: 32955516k total, 12861232k used, 20094284k free, 610956k buffers
    Swap: 2585588k total, 0k used, 2585588k free, 4194928k cached

    PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
    12123 root 20 0 679m 661m 1772 R 48 2.1 6:03.79 /usr/bin/unison
    12028 root 20 0 654m 636m 1772 R 46 2.0 8:23.31 /usr/bin/unison
    10339 root 20 0 675m 656m 1772 R 45 2.0 16:20.93 /usr/bin/unison
    10467 root 20 0 733m 709m 1772 R 45 2.2 13:18.29 /usr/bin/unison
    12222 root 20 0 650m 632m 1772 R 43 2.0 3:56.67 /usr/bin/unison
    10112 root 20 0 677m 659m 1772 R 42 2.0 22:25.94 /usr/bin/unison
    10206 root 20 0 690m 671m 1736 R 40 2.1 19:14.75 /usr/bin/unison
    12314 root 20 0 676m 657m 1736 R 39 2.0 1:53.87 /usr/bin/unison
    10558 root 20 0 683m 664m 1772 R 37 2.1 10:51.03 /usr/bin/unison
    10344 root 20 0 42152 6512 2216 S 2 0.0 0:32.48 ssh 192.168.1.131 -e none unison -server
    10115 root 20 0 42568 6940 2216 S 1 0.0 0:44.85 ssh 192.168.1.131 -e none unison -server
    10469 root 20 0 42856 7236 2216 S 1 0.0 0:25.40 ssh 192.168.1.131 -e none unison -server
    10562 root 20 0 42536 6908 2216 R 1 0.0 0:21.30 ssh 192.168.1.131 -e none unison -server
    12033 root 20 0 42088 6456 2216 S 1 0.0 0:16.18 ssh 192.168.1.131 -e none unison -server
    12126 root 20 0 41992 6372 2216 S 1 0.0 0:11.51 ssh 192.168.1.131 -e none unison -server
    12225 root 20 0 42792 6972 2216 S 1 0.0 0:06.97 ssh 192.168.1.131 -e none unison -server
    12318 root 20 0 42440 6808 2216 S 1 0.0 0:02.84 ssh 192.168.1.131 -e none unison -server
    10210 root 20 0 42344 6720 2216 S 1 0.0 0:39.14 ssh 192.168.1.131 -e none unison -server
    16 root 20 0 0 0 0 S 0 0.0 0:00.12 [events/1]

    please note, while load now is 8, its worked its way as high as 70 before killing things off and this has been rebooted just over 1.5 hours ago.

    about midnight or so either apache or mysql or postfix or all of them will die off till I run killall unison and then it starts it over again. The system is a dual xeon quad cores with 32 gigs of ram.. yet thats not enough.. crazy ...
     
  2. till

    till Super Moderator

    What might be is that the unison sync timespan is set too short, so that more then one unison instance is run by cron at a time. Change the cronjob on set it to a higher timespan. If this wont help, you might want to contact the unison developers and ask them what might cause this and check if there is a newer version of unison available which might fix the issue on your server.
     

Share This Page