LiteSpeed is changing the configuration

Discussion in 'Install/Configuration' started by justAndre, Mar 20, 2014.

  1. justAndre

    justAndre Member


    I have a strange problem with litespeed.

    If I let LS to use Cloudlinux with LVE in 5 days the configuration is changing to CageFS. I want to mention that nobody is accesing the server to change the configuration, I dont know why LS is making this change in the configuration file without a request from an administrator.

    I have exactly the same configuration on another server and there nothing happens with the settings.

    I dont want to use CageFS because all the sites are offline with this error on the admin panel:

    2014-03-20 08:03:46.415 ERROR [APVH_nexdddsr_Suphp:]: Failed to start one instance. Resouce limit reached!
    2014-03-20 08:03:46.416 ERROR [APVH_komparrts_Suphp:]: Failed to start one instance. Resouce limit reached!
    2014-03-20 08:03:46.417 ERROR [APVH_nexdddsr_Suphp:]: Failed to start one instance. Resouce limit reached!
    2014-03-20 08:03:46.417 ERROR [APVH_komparrts_Suphp:]: Failed to start one instance. Resouce limit reached!
  2. mistwang

    mistwang LiteSpeed Staff

    LS wont change configuration by itself. you should check file time stamp and admin console access log.
    Resouce limit is part of LVE, not CageFS, I guess you need to look into what and why resource limit has been reached.
  3. justAndre

    justAndre Member

    I'm not kidding, the configuration is changing without my intervention(i'm the only one who has the acces to LS admin interface). Today the configuration is with LVE and tomorow morning the configuration is on CageFS and all the sites are down due to the "Resouce limit reached" error.
  4. justAndre

    justAndre Member

    Hi again,

    Today litespeed crashed again, the error I see is this.

    Any idea what to do ?
  5. mistwang

    mistwang LiteSpeed Staff

    You need to raise LVE limit to allow two running instance of litespeed during a server graceful restart.
    If the limit has been reached by PHP processes started by the previous instance of litespeed, the newly started litespeed instance could not start any PHP process.

    Raise LVE limit to give some head room.
  6. justAndre

    justAndre Member

    Where do I need to change LVE limit? in cloudlinux or litespeed ?
  7. NiteWave

    NiteWave Administrator

    should be WHM -> Server Configuration -> CloudLinux LVE Manager
  8. justAndre

    justAndre Member

    I have incresed the pMEM to 512Mb
  9. NiteWave

    NiteWave Administrator

  10. justAndre

    justAndre Member

  11. NiteWave

    NiteWave Administrator

    I think EP is too low for a busy user.
    a lsphp5 process may fork more sub processes, for example, a simple php code
    exec("sh ls -l | head")
    will fork 2 additional sub processes

    anyway you can try to increase it to see if those errors disappear.
  12. sahostking

    sahostking Well-Known Member

    Did you solve this by any chance? We had 1 client report this today. Was watching apache logs then suddenly noticed all sites have this issue. When I disable CageFS all is good though but would like to have cagefs on again. Any ideaS?

    Note: Can't be setup issue as it hasn't changed in months. This is like magic :/
  13. NiteWave

    NiteWave Administrator

    you mean apache logs filled with
    "Failed to start one instance. Resouce limit reached!" ?

    per discussion above, it's likely a configuration issue : LVE limit too low.
    if the issue happened only once, it's hard to find out the root cause. maybe just a restart fix it.
    if it happens repeatedly, try to raise LVE limit as suggested, see if any improvement.

    also, you can try latest 4.2.19. to be sure to run tests against latest version.

Share This Page