Apinger.status constantly updated?



  • Hi all,

    Trying to minimize as much disk write as possible in a VMware environment. While watching "top -a"/"m" I kept seeing this:

    
      PID USERNAME   VCSW  IVCSW   READ  WRITE  FAULT  TOTAL PERCENT COMMAND
    63884 root          8      6      0      1      0      1 100.00% /usr/local/sbin/apinger -c /var/etc/apinger.conf
    
    

    .. and tracked it down to a write to /tmp/apinger.status every 3 or so seconds. Is there a way, without running an embedded install (I use packages sometimes), to move this to memory? Perhaps move /tmp to a ramdisk? apinger config?

    Thanks!
    Ben

    edit
    I should note, I did find /var/etc/apinger.conf which specifies the file and the time to write, but I wasn't sure if modifying that would a) make any difference since it seems like the file is dynamically created by something else since my IP is in there and b) if changing the write time/file would also change apinger's interval of checking connections.


  • Rebel Alliance Developer Netgate

    Just edit /etc/inc/gwlb.inc and find all the references to apinger.status (there are only ~5) and change it to /var/run/apinger.status



  • @jimp:

    Just edit /etc/inc/gwlb.inc and find all the references to apinger.status (there are only ~5) and change it to /var/run/apinger.status

    Thanks, updated to the variable for varrun_path on those lines, but another question comes to mind. For something that is updated every 5 seconds, is there a reason it doesn't default to a memory based file system?

    Also, is my system set up right? I thought /var/run would be a tmpfs, but it's a ufs?

    /dev/da0s1a on / (ufs, local)
    devfs on /dev (devfs, local)
    /dev/md0 on /var/run (ufs, local)
    

    Still seeing IO writes, which is what makes me curious.

    Thanks,
    Ben


  • Rebel Alliance Developer Netgate

    What type of install are you on? And what version?

    Even my full installs have /var/run as a memory disk



  • @jimp:

    What type of install are you on? And what version?

    Even my full installs have /var/run as a memory disk

    2.0-RELEASE (i386)
    built on Tue Sep 13 17:00:00 EDT 2011

    Regular hard drive install, though in a VMware session. What does a "mount" look like for your system?


  • Rebel Alliance Developer Netgate

    Here's a few different ones:

    : mount
    /dev/ad4s1a on / (ufs, local)
    devfs on /dev (devfs, local)
    /dev/md0 on /var/run (ufs, local)
    devfs on /var/dhcpd/dev (devfs, local)
    
    : mount
    /dev/ad0s1a on / (ufs, local)
    devfs on /dev (devfs, local)
    /dev/md0 on /var/run (ufs, local)
    procfs on /proc (procfs, local)
    devfs on /var/dhcpd/dev (devfs, local)
    
    : mount
    /dev/ad0s1a on / (ufs, local)
    devfs on /dev (devfs, local)
    /dev/md0 on /var/run (ufs, local)
    devfs on /var/dhcpd/dev (devfs, local)
    


  • @jimp:

    Here's a few different ones:

    Those /var/run's look just like mine. What identifies them as a tmpfs? (Sorry, used to Fedora/RedHat/etc, not FreeBSD, and unsure why top would report IO if it's in ram).

    Thanks,
    Ben


  • Rebel Alliance Developer Netgate

    /dev/md0 is a memory disk.



  • @jimp:

    /dev/md0 is a memory disk.

    Haha, ahh, thanks. I guess it just reports IO even if it's a memory disk, then.

    Thanks,
    Ben


Locked