Kaseya Community

Disk Full error at month-end

  • I am using 250Gb external USB hard drives. This particular server has approx. 86Gb to backup. The initial tib file is about 55Gb and a month's worth of incremental backups was 60Gb. I have it set to 2 month retention on a 1 month cycle.



    Today was the 30th day and it created new folders and base tib files. It worked fine on two of the three servers, but this server had a failure. I looked at the log on the Kaseya console and it said Failed - Disk full. Here is what I believe happened:



    Apparently, the backup process checks to see if 30 days have passed. If so, it does a new base backup in a new folder and then deletes the folder from the backup a month prior to the previous month. However, this server has a 55Gb base tib and if you add up two previous months worth of backups and a third base tib file before it deletes the previous month, it exceeded the capacity of the 250Gb drive.



    Now (after the failure),the external drive is sitting there with an empty\servername.groupname\VolBackup\20061030 folder and the previous August folder has been deleted. Since the prior month (or prior period) folder is going to end up deleted no matter if the backup was successful or not, could the order of things in the way Kaseya does the backup, change slightly? Could it delete the prior month before attempting to do a new base tib?



    I know that eventually 250GB will not be sufficient, but this issue seems to have a simple solution with a very benign effect and cansave having to upgrade drives for quite a while.



    Thanks,



    Randy Spangler

    MERITSolutions


    Legacy Forum Name: Disk Full error at month-end,
    Legacy Posted By Username: JustinWC
  • One more twist to report...

    As I stated, we have three servers in this group that are all on the same schedule. Also, remember that one of the backups failed to create the new initial tib file this month.

    Well, last night it created the new one as expected, however it created a new folder, 20061031 01.00.41, to hold the new file. Not what I thought would happen, because it now looks like one of the backups will be a day out of sync with the other ones. Not fatal, but kinda messes thing up a bit.

    What else happened was also unexpected. When I looked in the new 20061030 folder for the other two servers, I found that the initial base tib file from yesterday was replaced with a new, large tib file dated today! Not another numbered incremental tib file, but a new base tib file. Did this have anything to do with the failure of 1 of 3, or did it have something to do with being on a 30 day rotation and today being the 31st of the month? Of course, this totally hosed my scheme to copy the big tib file out to a non-replicatingfolder while the numbered tibs were sent to the offsite storage facility for the rest of the month.

    The law of unintended consequences RULES!!!

    Randy

    (BTW, my first posting in this thread was done under the cached logon of Justin, instead of myself. Sorry for the confusion...)


    Legacy Forum Name: Backup Tab,
    Legacy Posted By Username: RandyS
  • Well, here we go again... I moved the big tib files out into a non-replicated folder and guess what? It creates a new one. Perhaps it is because the folder is empty and there is no numbered tib to use as a reference to say that the sequence is proceeding.

    Randy


    Legacy Forum Name: Backup Tab,
    Legacy Posted By Username: RandyS
  • Incremental backups capture all changes from the full backup tib file. If the full backup tib has been moved out then the incremental is capturing all changes relative to zero (basically a new full backup).

    Separating the full from the incremental files is generally not a good idea anyway because all the files will be needed to restore or mount the volume anyway.


    Legacy Forum Name: Backup Tab,
    Legacy Posted By Username: Mark Sutherland