Time Machine "Event store UUIDs don't match for volume" after swapping disk

10,074

Solution 1

To fix this: wait.

  • After performing a full restore, Time Machine will always create a full backup, by design. Without knowing why Apple thinks this is required, I'd favor a reliable backup over time and disk space. See also Apple's Mac OS X 10.5: Time Machine performs full backup after a full restore.

  • In all other cases: Time Machine has detected that it cannot tell what's on your backup, and what's not, and needs to compare both. You're probably also seeing Node requires deep traversal.

This is not related to the ID of the disk (the hardware) itself. TM keeps the FSEvents ID it used for the last backup in the "extended attribute" com.apple.backupd.SnapshotVolumeLastFSEventID on the disk. Normally, all it takes to determine what has changed is to compare that value to the ID known by OS X. However, if for some reason the OS X FSEvents database can no longer be trusted, it creates a new one, which changes its unique UUID. TM checks to see if the FSEvents database can be used for a specific backup disk by comparing that unique UUID to the UUID that is stored with the backup, in com.apple.backupd.SnapshotVolumeFSEventStoreUUID. So after a new FSEvents database is created, these UUIDs no longer match and TM needs to compare the harddisk with the backup, or might need to create a full backup.

Solution 2

I found that the UUID issue is resolved and the backup continues after many messages like this:

3/15/12 1:49:35.010 PM com.apple.backupd: Waiting for index to be ready (100)

BUT, only IF there is enough space on the backup drive. My drive was very full, and this 'waiting for index' lasted forever until I reclaimed some space on the drive.

Share:
10,074

Related videos on Youtube

Jules
Author by

Jules

Updated on September 18, 2022

Comments

  • Jules
    Jules over 1 year

    My new hard drive died the last week and had to put my old drive backup into my Mac Mini, which is running Snow Leopard. I was then able to restore my latest Time Machine backup.

    When I upgraded a few months ago I used Carbon Copy and I had permission problems.

    So I have my old drive in my system at the moment, but when I try to do a Time Machine backup, it's VERY slow. It's using the same settings / locations as before. I download TM Buddy, which says...

    Starting standard backup
    Backing up to: /Volumes/Mac Time Machine/Backups.backupdb
    Event store UUIDs don't match for volume: Macintosh HD
    Waiting for index to be ready (100)
    Waiting for index to be ready (100)
    Node requires deep traversal:/ reason:must scan subdirs|new event db|
    No pre-backup thinning needed: 109.39 GB requested 
          (including padding), 121.15 GB available
    

    I'm trying to do a backup so I can put in another new drive, so I can do a Time Machine restore, like I did last week.

    What can I do to fix this problem?

    • slhck
      slhck over 12 years
      The problem is that your Mac's disk UUID doesn't match the one stored on the TM volume. You are probably creating a new backup instead of incrementally updating the old one. The solution would involve changing the UUID on the TM volume. I have to look if I can gather some information, should be solvable.
    • Jules
      Jules over 12 years
      That makes sense, look forward to your solution. I did try this web.me.com/pondini/Time_Machine/A4.html which deletes com.apple.TimeMachine.plist which hasn't helped.
    • Jules
      Jules over 12 years
      It also says Backing up 32kb of 92.25GB after about 20 minutes :(
    • Arjan
      Arjan over 12 years
      And was this indeed full restore to the old disk, hence completely wiping out the old contents of that old disk? (In other words: just like if it were a totally different disk?)
    • Jules
      Jules over 12 years
      Yes it was a full restore using the option from the setup program from my grey dvd.
    • Jules
      Jules over 12 years
      It was my original disk that I restored to, the disk provided my with my mac mini, with snow leopard pre-installed
    • Arjan
      Arjan over 12 years
      @Jules, I assume the full restore wiped anything that was on that old disk. Anyway: your edit clearly shows new event db, indicating that OS X either had no FSEvents database at all after the restore (makes sense) or somehow invalidated it itself because it knew a restore might have messed up with its state. I really think you'll have to wait...
    • Jules
      Jules over 12 years
      OK, whats your best guess as to how long it will take ?
    • slhck
      slhck over 12 years
      That may range from a couple of hours to one day (from what I've experienced and read).
    • Arjan
      Arjan over 12 years
      92GB made me think it was an incremental backup. But: how much data is on your harddisk?
    • Jules
      Jules over 12 years
      The hard disk is 120gb and about 100gb used.
    • slhck
      slhck over 12 years
      Well, 109.39 GB were requested, and 92 GB after thinning, that should be fine.
    • Arjan
      Arjan over 12 years
      Ah, that seems like a full backup after all then. If you want to avoid that (but: I guess you might NOT want to avoid it, as you want a good backup to restore to the new disk that you're about to install; waiting now might be more secure...?), see the answer @slhck posted.
    • Jules
      Jules over 12 years
      Ooooo it just shot up to 169mb of 95.29gb :)
    • slhck
      slhck over 12 years
      Generally, give it time. I've had my issues with Time Machine as well, and it sort of fixed itself eventually.
    • Jules
      Jules over 12 years
      Copied 42 KB of 88.7 GB, 61 of 340988 items, Copied 161.7 MB of 88.7 GB, 11926 of 340988 items
    • Arjan
      Arjan over 12 years
      Did this meanwhile complete? ;-)
  • Jules
    Jules over 12 years
    In fact I started the backup this afternoon and after an hour it had done 52kb. Surely there must be something I can do, or is it lying to me?
  • Arjan
    Arjan over 12 years
    TM only knows if something changed for each directory. That 92.25GB might be the total size of all changed directories, but the files that have actually changed might be smaller in total. So maybe the 35kb is not really 35kb of 92.25GB. Don't know.
  • slhck
    slhck over 12 years
    Are you sure it's not the issue in my answer (currently deleted below)? Jules changed their disk, therefore the UUIDs won't match.
  • Jules
    Jules over 12 years
    I've added the full text from TM Buddy to my question
  • Arjan
    Arjan over 12 years
    (That was a nice answer, by the way, @slhck. But indeed I think it does not apply, but I admit I had to read the question twice before answering, to get an idea of the events...)
  • slhck
    slhck over 12 years
    I see, you have a point. Maybe I'm confused by the fact that I've had similar issues with different partitions before. "Node requires deep traversal" usually means that things will sort out eventually.
  • slhck
    slhck over 12 years
    By the way: My answer found a new home in case you'd like to quickly peer review it!
  • Arjan
    Arjan over 12 years
    Be sure to read that, @Jules. I might have been wrong. (Maybe waiting just gets you a full new backup, rather than an incremental backup!)
  • Jules
    Jules over 12 years
    Right, erm, can I do that now, confused ?
  • Arjan
    Arjan over 12 years
    I don't what to advice you, @Jules. Waiting won't hurt. The other trick might speed up things. But if Time Machine feels it cannot safely use the old backup as a base for the backup of your restored Mac, then maybe that's for a good reason. See also Apple's Mac OS X 10.5: Time Machine performs full backup after a full restore.