Ad Widget

Collapse

Too many processes busy

Collapse
X
 
  • Time
  • Show
Clear All
new posts
  • stepassia
    Junior Member
    • Feb 2018
    • 27

    #1

    Too many processes busy

    Hi to all, I've a problema with my zabbix server.
    Several times per day (4 or 5) zabbix fire the busy triggers (like Zabbix history syncer processes more than 75% busy or Zabbix alert syncer processes more than 75% busy) and, at the same time, any zabbix agent can't be reached (with hundred alert mail).

    This my zabbix parameter:
    • Zabbix server is running Yes localhost:10051
    • Number of hosts (enabled/disabled) 126 108 / 18
    • Number of templates 85
    • Number of items (enabled/disabled/not supported) 13932 9583 / 3471 / 878
    • Number of triggers (enabled/disabled [problem/ok]) 6627 2886 / 3741 [12 / 2874]
    • Number of users (online) 16 3
    • Required server performance, new values per second 65.46

    Virtual Server detail:
    • Database: mysql on localhost
    • RAM: 3.5Gb
    • CPU: 2
    And the Zabbix Server Configuration:
    • StartPollers=30
    • StartPollersUnreachable=50
    • StartPingers=15
    • StartDiscoverers=3
    • StartAlerters=5
    • VMwareCacheSize=64M
    • HousekeepingFrequency=1
    • MaxHousekeeperDelete=0
    • CacheSize=256M
    • HistoryCacheSize=1024M
    • HistoryIndexCacheSize=128M
    • TrendCacheSize=256M
    • ValueCacheSize=256M
    • Timeout=5
    • LogSlowQueries=1500
    Any idea to solve this issue?
    Thanks a lot!
    Stefano
  • stepassia
    Junior Member
    • Feb 2018
    • 27

    #2
    Thank you Cyber!

    Comment

    • cyber
      Senior Member
      Zabbix Certified SpecialistZabbix Certified Professional
      • Dec 2006
      • 4807

      #3
      I think you have unreasonably many Pollers and pingers for such small amount of hosts and items.

      MaxHousekeeperDelete=0 ... docs say about this "If set to 0 then no limit is used at all. In this case you must know what you are doing, so as not to overload the database! 2"

      Comment

      • cyber
        Senior Member
        Zabbix Certified SpecialistZabbix Certified Professional
        • Dec 2006
        • 4807

        #4
        Then maybe other paramters of your servers are a small, if possible add some memory and cpu. Considering you have DB on same host as server, it may be a bit too few....

        Comment

        • cyber
          Senior Member
          Zabbix Certified SpecialistZabbix Certified Professional
          • Dec 2006
          • 4807

          #5
          Its just my gut feeling, so don't take it as a fact, but I think alerter, http poller and LLD messages are the result and history syncer is the reason... I can see similar behaviour sometimes in our environment also... if in some reason history syncer is busy, at same time LLD queue goes up... History syncer needs quick DB. But I am not too good on any of this platform stuff.. I do not have any suggestions for you here...

          Comment

          Working...