OpenVZ Forum: Support » user_beancounters: failcnt increasing before resources reach limit

Home » General » Support » user_beancounters: failcnt increasing before resources reach limit

Show: Today's Messages :: Show Polls :: Message Navigator
E-mail to friend

user_beancounters: failcnt increasing before resources reach limit [message #49131]

Thu, 28 March 2013 12:29

david2389
Messages: 1
Registered: March 2013

Junior Member

Hi,

I'm a customer of a VPS provider who uses OpenVZ. I don't have access to the OpenVZ configuration or kernel logs.

I'm finding those typical problems reported elsewhere (e.g. "fork: Cannot allocate memory"). I understand they happen because I'm using too much resources. However, the "maxheld" value in the /proc/user_beancounters file isn't near the "barrier" or "limit" ones, at which point I'd expect the "failcnt" to increase.

To illustrate my problem, consider this partial user_beancounters file:

Version: 2.5
       uid  resource                     held              maxheld              barrier                limit              failcnt
 67767067:  kmemsize                 26743189             26895615             31457280             34603008              9201127
            privvmpages                473723               485794              1048576              1153433                30530
            numproc                        87                   89                  128                  128                  635

Some minutes later, this is how it looks:

Version: 2.5
       uid  resource                     held              maxheld              barrier                limit              failcnt
 67767067:  kmemsize                 26960203             28481774             31457280             34603008              9227898
            privvmpages                431904               599424              1048576              1153433                33525
            numproc                        88                   90                  128                  128                  635

We can see that the failcnt for kmemsize and privvmpages has increased, however the maxheld field doesn't seem to have hit the limit for each field. I'd expect to find something like this:

       uid  resource                     held              maxheld              barrier                limit              failcnt
 67767067:  kmemsize                 26960203             34603008             31457280             34603008              9227898

Is this normal behaviour? Is there something I can do to make a better usage of the resources I've been allocated?

Report message to a moderator