files sometimes back up, sometimes fail
  • paleolith
    0
    I've created a local backup plan and it's run three times. I'm having some trouble understanding the messages.

    On the first run, the email report says "Error: Warning.[sic!] One or more backup paths don't exist". Later it says "files failed to backup: 3". Does this mean that the "failure to backup" is the "warning"? (I can believe that an occasional file gets deleted before it gets backed up. That's not a problem; I'm just trying to understand the messages.)

    The more bothersome one, though, is the first update run -- which ran immediately after the full backup completed, as it was already overdue. The email report says "Error: Cannot access 53 files because they are being used by another process" and "Files failed to backup: 53". The files listed are unsurprising: SQLite, parent.lock, that kind of stuff. Actually mostly files I don't want to back up anyway. Looking in the backup directory, it seems all these files were backed up on the full backup, and were backed up again on the SECOND update run. So why did they all fail on the first update run?

    Thanks,

    Edward
    paleolith Options
  • Matt
    91
    ↪paleolith

    On the first run, the email report says "Error: Warning.[sic!] One or more backup paths don't exist". Later it says "files failed to backup: 3". Does this mean that the "failure to backup" is the "warning"? (I can believe that an occasional file gets deleted before it gets backed up. That's not a problem; I'm just trying to understand the messages.)
    Yes, that's just a warning message about some files that are failed, so most of your data except those 3 files is safely placed on storage side.

    Regarding those 53 failed files: Most likely they were actually in use by some other process during the backup. I suggest to make sure "Always use VSS" and "Use system VSS provider" are enabled for that plan to avoid issues like that.
    Matt Options
  • paleolith
    0
    Oh, I'm sure they were in use. Many of them are Firefox and Chrome files in the AppData folder -- SQLite files. The question is not why they failed on backup #2, but why they did NOT fail on backups #1 and #3 (and as of today #4). Or IOW why the discrepancy in behavior under what appears to me to be the same conditions.

    Thanks for the advice on VSS; I'll do that. It was not clear to me from the docs whether this is an expensive option; I gather you are saying it's not.

    Thanks,

    Edward
    paleolith Options
  • Matt
    91
    ↪paleolith
    Local files that locked by other software are impossible to back up without VSS, so currently my only assumption is that they were actually not in use by the time you were launching your second run of the plan.
    Matt Options
  • paleolith
    0
    ↪Matt
    Oh, I guess I was confused by the word "force" -- I was assuming that with the option reset, VSS would be used when needed, but with the option set, always. Is there a reason the option isn't just called "Use VSS"?

    Thanks,

    Edward
    paleolith Options
  • Matt
    91
    ↪paleolith
    This is just how our developers decided to name that option. It simply switches VSS on and off during backups.
    Matt Options
  • Craig
    0
    Good Afternoon this is my first post but I am currently experiencing a similar issue where our SQL daily backup job is failing stating that one or more backup paths don't exist however this will continue for a few days. Strangely enough if we clone the backup job and start the schedule then the backups complete successfully for several weeks. Support pointed me to local credentials but cloning the job resolves the issue all be it temporarily. On a side note we had to disable the VSS because this was causing a lot of issues with our SQL performance.

    Any help would be greatly appreciated. I have inherited cloudberry from my development team so now getting to deal with it in the wild so to speak.

    thanks again in advance

    Craig
    Craig Options
  • Matt
    91
    ↪Craig
    I only see the original logs and the error message that refers to network credentials.
    I suggest to re-send the logs to the original ticket after making sure the credentials are 100% correct and continue communication there. Also, since it's a hybrid backup I would suggest to synchronize the repository for both storage accounts via tools > options > repository, but the first thing is to make sure the software can transfer the data to the local destination.
    Matt Options
  • Craig
    0
    HI Matt can you message me the ticket number as it was my colleague who raised it initially
    Craig Options
  • Matt
    91
    ↪Craig
    Forwarded my colleague's latest reply to your address(that's the only way, I guess).
    Ticket number should be #86554

    Just noticed that there's also ticket #101882, which is newer, but the problem is similar. You can also request to add you to that ticket directly via
    Matt Options
bold
italic
underline
strike
code
quote
ulist
image
url
mention
reveal
youtube
tweet
Add a Comment