1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.

processes dying

Discussion in 'Tomato Firmware' started by gid_, Dec 15, 2010.

  1. gid_

    gid_ Addicted to LI Member

    Processes have been dying on my router recently. Such as telnetd, syslogd, klogd, etc. Other seems to stop responding, such as dnsmasq.

    I recently upgraded to 1.28 from 1.23, but I can't remember if I upgraded for the heck of it, or if I upgraded to hopefully help solve this issue--I'm thinking the later, but I'm not sure.

    Anyone have any ideas? Hardware dying maybe? Are there other things to check out? /var/log/messages seems pretty sparse, at least when syslogd happens to be running. :)

    WRT54GL
     
  2. mstombs

    mstombs Network Guru Member

    It is always recommended you reset to defaults and re-type your config when strange things like this happen. BUT you do not normally need to do this for simple version increments of stock Tomato. It is not typical of Tomato 1.28 - I have a copy running 10 days since I last rebooted, on a router that's seen every version since about 1.07 (and HyperWRT Thibor before that).

    Tasks do get killed by kernel when router runs out of memory - often dnsmasq because it is big and runs as restricted user nobody. Dnsmasq gets restarted by default, unless you have changed that config item. Do you have any strange scripts loaded? There was a case where a router brought down by multiple copies of looping script.
     
  3. gid_

    gid_ Addicted to LI Member

    I only have an extra config line for dnsmasq, addn-hosts=/cifs1/hosts.

    I have QoS enabled, UPnP, and cifs 1 enabled, and maybe a dozen or so forwards. The only thing I've really done to router in the past year is upgrade the firmware and add the addn-hosts line to dnsmasq.

    It looks like I've been running tomato since v1.13 or so, or at least that's the earliest version on my HD.

    I don't suppose there's anyway to convince linux to not kill some processes?
     
  4. gid_

    gid_ Addicted to LI Member

    If it is ram related, would trimming down my QoS rules help maybe? Or any other options I can pass to dnsmasq so it won't be a hog.

    Alternatively, does anyone have experience with the Netgear WNR3500L-100NAS and Tomato?
     
  5. mstombs

    mstombs Network Guru Member

    This could be a reason, suggest try without. if that file not available when dnsmasq starts it likely aborts - and dnsmasq always get restarted after every wan reconnection and many other config changes.
     
  6. gid_

    gid_ Addicted to LI Member

    Thanks, I'll give that a try. I can just add to the existing hosts file in a startup script.
     
  7. gid_

    gid_ Addicted to LI Member

    Processes are still dying after removing the addn-hosts line for dnsmasq. Any other ideas short of a reset? How can I find out what process is using up too much memory? Nothing seems amiss right now, here's the ps output:


    # ps
    PID USER VSZ STAT COMMAND
    1 root 1724 S init noinitrd
    2 root 0 SW [keventd]
    3 root 0 RWN [ksoftirqd_CPU0]
    4 root 0 SW [kswapd]
    5 root 0 SW [bdflush]
    6 root 0 SW [kupdated]
    7 root 0 SW [mtdblockd]
    26 root 1700 S buttons
    72 root 1932 S klogd
    84 root 1724 S nas /etc/nas.conf /var/run/nas.pid lan
    92 root 1960 S crond -l 9
    99 root 1260 S rstats
    102 root 0 SW [cifsoplockd]
    103 root 0 SW [cifsd]
    107 root 1652 S httpd
    162 root 1000 S miniupnpd -f /etc/upnp/config
    285 root 1952 S udhcpc -i vlan1 -s dhcpc-event -H router -S
    451 nobody 864 S dnsmasq
    483 root 1940 S telnetd -p 23
    511 root 1972 S -sh
    515 root 1948 R ps
     

Share This Page