Pfsense 2.01 on Virtualbox: disk full, DHCP stops working



  • Hello -

    I have been running pfsense on virtualbox for quite some time and it works well, but every now and then DHCPD dies, it says:

    dhcpd: Can't create new lease file: No space left on device

    At this point I can't log into the web page, but the console still works.

    df -i brings up:

    
    Filesystem  1K-blocks   Used    Avail Capacity iused   ifree %iused  Mounted on
    /dev/ad0s1a  19295222 261306 17490300     1%    9418 2487092    0%   /
    devfs               1      1        0   100%       0       0  100%   /dev
    /dev/md0         3694     48     3352     1%      25     741    3%   /var/run
    devfs               1      1        0   100%       0       0  100%   /var/dhcpd/dev
    
    

    This vm is running on pretty new hardware, an HP DL380 G7 running RAID 5 with plenty of real hd space available. I don't think it's hardware failure. I have the virtual disk set to expand as necessary. Would it help to allocate all the disk space when I build the vm? Are there any tricks to running Pfsense on virtual box that would help here?

    Thanks!


  • Rebel Alliance Developer Netgate

    What (if any) errors show on the console? That error can either mean it's out of disk space (which it doesn't appear to be) or that the disk disappeared or had problems.



  • "dhcpd: Can't create new lease file: No space left on device" is one error I see on the console. Since it's not happening now, I will have to wait until the next time to catch more. Also, I can't log into the web configurator - there are php errors at the top of the login page. I will try to capture more the next time it happens, but when it's down, my phone rings off the hook, so there's not a lot of time to fix it.

    Like I said, the hardware is nearly new. The virtual disk is not full as shown by df but it sure acts like it. I cranked up a new iteration of this firewall and told VirtualBox to allocate all the disk space up front. Let's see if that makes any difference.



  • If i remember correctly there was same kind of error earlier and that was caused because /var was running all the space it had mounted.
    And you df -i shows the same thing /var %iused is 100%


  • Rebel Alliance Developer Netgate

    His df only shows iused 100% on */dev/ filesystems. This is normal.



  • After rebooting the machine at 6:00am this morning, now at 3:00pm I had to reboot. The error on the web login screen looks like this:

    "Warning: fopen(/tmp/config.lock): failed to open stream: Device not configured in /etc/inc/util.inc on line 123 Warning: flock() expects parameter 1 to be resource, null given in /etc/inc/util.inc on line 134 Warning: fclose(): supplied argument is not a valid stream resource in /etc/inc/util.inc on line 135 Warning: session_start(): open(/var/tmp//sess_4afee71a74f8b38efb6da851082224e1, O_RDWR) failed: No space left on device (28) in /etc/inc/auth.inc on line 1254"

    and on the console it says,

    /etc/rc.initial: /bin/tcsh: Device not configured
    /: create/symlink failed, no inodes free
    /: write failed, filesystem is full
    Fatal error: Unknown: apc_fcntl_lock failed: in Inknown on line 0

    So it's acting like the drive is full. The virtual hardrive is 30gb and I have these packages installed: Cron, mailreport, NRPE v2, and widescreen.


  • Rebel Alliance Developer Netgate

    Hardware failures for storage also present as a "full" disk because they both result in a failed write.

    If the disk is not full, it can't be reached.



  • Well, it could be some sort of virtual hardware failure, but the real hardware is just fine.


  • Rebel Alliance Developer Netgate

    It's possible, yes, but either way it's hardware (real or virtual) and nothing pfSense can do there.

    You might try a different storage controller type in virtualbox.



  • Does pfsense like one better than another? I'm using IDE right now. The choices are IDE, SATA. SAS and SCSI.


Locked