Skip to main content

CRON Error Cannot allocate memory

Mark Coates
Hi all this morning i recieved this email about a cron job which i was unaware ran as my emails are normally overnight. this is the content of the email; From: root@xxxxxxxxx.co.uk Subject: Cron /usr/bin/test -e /etc/cpanel-dovecot-solrdisable || /usr/local/cpanel/3rdparty/scripts/cpanel_dovecot_solr_commit Content: /bin/sh: fork: Cannot allocate memory
i have not seen this before and should it be something to be concerned about? i have added xxxxx to hide the domain.

Comments

7 comments

Date Votes
  • nixuser
    There might be a possibility that the free memory exhausted while running the script. Is it a vps?
    0
  • Mark Coates
    Looking on munin there was plenty of memory available. Yes it"s a vps
    0
  • cPanelLauren
    Is this possibly a VPS using Virtuozzo? If it is, what's the ouptut of the following: cat /proc/user_beancounters
    The error indicates that the system had no free memory at the time the cron was run (this wouldn't be attributed to anything else)
    0
  • Mark Coates
    Is this possibly a VPS using Virtuozzo? If it is, what's the ouptut of the following: yes it is and here you go :) cat /proc/user_beancounters
    The error indicates that the system had no free memory at the time the cron was run (this wouldn't be attributed to anything else)

    Version: 2.5 uid resource held maxheld barrier limit failcnt 75533: kmemsize 60754625 73871360 67158736 73874609 4034 lockedpages 0 3279 3279 3279 382 privvmpages 1152530 1310720 1310720 1310720 116426 shmpages 75572 362264 9223372036854775807 9223372036854775807 0 dummy 0 0 9223372036854775807 9223372036854775807 0 numproc 310 429 1000 1000 0 physpages 1301369 3740721 0 9223372036854775807 0 vmguarpages 0 0 130863 9223372036854775807 0 oomguarpages 712272 835153 130863 9223372036854775807 0 numtcpsock 73 234 5000 5000 0 numflock 692 866 5000 5000 0 numpty 1 5 163 163 0 numsiginfo 0 99 1024 1024 0 tcpsndbuf 1803256 8912632 15676997 22386245 0 tcprcvbuf 1288984 13718272 15676997 22386245 0 othersockbuf 457776 2847208 7838498 14547746 0 dgramrcvbuf 0 283400 7838498 7838498 0 numothersock 250 545 1638 1638 0 dcachesize 15095353 15095808 14656124 15095808 0 numfile 4554 7350 52416 52416 0 dummy 0 0 9223372036854775807 9223372036854775807 0 dummy 0 0 9223372036854775807 9223372036854775807 0 dummy 0 0 9223372036854775807 9223372036854775807 0 numiptent 300 1000 1000 1000 2194 [root@dedivps-75533 ~]#
    0
  • cPanelLauren
    This is showing you've exceeded the resources available multiple times based on the failcnt - so your allocated resources may not have been exceeded but the physical resources weren't available on the node. You can see this shown in the failcnt column - The field failcnt shows the number of refused "resource allocations" for the whole lifetime of the process group. You can read more about how this all works here: /proc/user_beancounters - OpenVZ Virtuozzo Containers Wiki UBC secondary parameters - OpenVZ Virtuozzo Containers Wiki UBC auxiliary parameters - OpenVZ Virtuozzo Containers Wiki But ultimately it sounds like the resources were definitely not available and you may want to discuss this further with the provider.
    0
  • Mark Coates
    I might be acting thick on this one but if the memory is fully used should this then use the SWAP as a temporary measure?
    0
  • cPanelLauren
    Not if the resources aren't available on the node, which is what appears to be occurring here. Your container is requesting resources and it's failing to obtain them from the hostnode.
    0

Please sign in to leave a comment.

Didn't find what you were looking for?

New post