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

One big problem I notice with WHM transfer.

Discussion in 'General Discussion' started by AdamDresch, Apr 14, 2011.

  1. AdamDresch

    AdamDresch Member

    Joined:
    Jun 22, 2006
    Messages:
    78
    Likes Received:
    0
    Trophy Points:
    6
    I noticed this on the last server I moved from, and again tonight when I tried to move 20 odd accounts.

    The time out value in ssh, by default, I think is 2? minutes, so if the pkgacct process takes longer than 2 minutes, it times out.

    The pkgacct process carries on on the source server and the destination server restarts the process again, thus, starting yet another pkgacct process.

    Tonight I noticed there were 5 pkgacct processes for the same (largish) domain, which actually ended up killing my source server.

    I'm still waiting on BurstNet to find out what's gone wrong, but it's been inaccessible for an hour now, heh.
     
  2. cPanelTristan

    cPanelTristan Active Member
    Staff Member

    Joined:
    Oct 2, 2010
    Messages:
    7,621
    Likes Received:
    4
    Trophy Points:
    38
    Location:
    somewhere over the rainbow
    If you have this type of issue happen again, please open up a ticket if you have root SSH to the machine. If you do not, please have your provider open up a ticket with us. This way we can investigate precisely what is occurring.

    Personally, I've never seen the process restarted after 2 minutes, since I'm on the migration team as one of the members and we transfer a large number of accounts from other control panels to cPanel. Oftentimes, those accounts are very big. Now, I've seen SSH timeout due to a network disconnect, and I've also seen it happen when there's a very long time delay (let's say 2 hours or something on one account), but never on this short of an interval, since almost always the pkgacct process is going to take longer than 2 minutes.

    Also, when you performed the transfers, which specific option in WHM > Transfers area was chosen?
     
  3. AdamDresch

    AdamDresch Member

    Joined:
    Jun 22, 2006
    Messages:
    78
    Likes Received:
    0
    Trophy Points:
    6
    Hiya, I'll get back to you on that when/if the server comes back up

    From memory, I chose the default options for transfering, using the root login

    It starts off ok, then gets to the bit where it says something like Copying maildir/www folder, at which point a pkgacct process seems to start on the source server

    Because it's taking so long to package it up, the destination transfer says something about an SSH/SSL? time out, then it restarts the process, but in doing so, it starts another pkgacct process on the source server, even though one is already running, so it just ends up bogging the server down.

    In my case, I was unable to terminate pkgacct processes, so it just seemed to crash my server heh.
     
  4. AdamDresch

    AdamDresch Member

    Joined:
    Jun 22, 2006
    Messages:
    78
    Likes Received:
    0
    Trophy Points:
    6
    Well, not sure what's happened, but the kernels are all missing.

    It appears that all of the kernels were removed from this server. While the grub prompt still has our BurstNET kernel listed, trying to boot it reports file not found. At this point the only option would be to reinstall the server. The standard fee for this service is $25.00 USD. This fee is waived for reinstalls due to hardware failure.

    Ah well.
     

Share This Page