A bug in the OS X 10.5.3 update creates trust issues with the reliability of some Time Machine backups, writes Baltimore Sun reporter David Zeiler. Hourly system backups to some Mac Pro machines are inconsistently met with the vague error message
which leaves the option of staring at the screen or clicking the OK button and pretending the failed backup doesn’t matter.
MacRumors has had a discussion thread going on this topic since the end of May, and the support forums on the Apple website show a question on this topic that remains unanswered after 69 replies.
A simple fix may help in some cases, according to blogger David Alison. Run the Console application in your Utilities folder, and select All Messages on the left. Then start searching using the box in the upper right. All Time Machine activity is logged under the process name of “backupd”, so searching for that will pull up all the relevant logs. If you’ve got an open backup that’s listed as “In Progress,” even though Time Machine is not running, try deleting that to see if it allows your backups to continue.
14 responses to “Time Machine Bug Raises Backup Reliability Questions”
I’ve been getting that message lately with my time machine back up, but I haven’t had time to even think about it. I guess I should be worried. I’ll try the fixes. Thanks.
I get that on my G5 iMac. Going to TM’s console and choosing backup now seems to solve the immediate issue but it would be nice not to have to do this manually from time to time.
The most common error message I get is that some random .DS_Store file can’t be found. Was I under the misapprehension that .DS_Store files should stay put once they’re created? Fortunately these errors clear up next time Time Machine runs.
I also get more serious meltdown-class errors when my TM disk is close to full. I have been solving these by deleting blocks of my backup where incremental data in unimportant to me, thus freeing up disk space until the next meltdown.
I have been getting errors like this one, even though I go on with a hard core backup strategy, meaning copying all files and checking the total number of items and sizes via info manually. Only this way I have the chance to dig through to find those files, which cannot be copied due to non ANSI encoded name or ill permissions. By the way, I had experienced no problems with this strategy in the days of Tiger. So I rather assume that the “bug in the OS X 10.5.3” is related to file copying mechanism rather than the fancy Time Machine backing-up.
I get this regularly on my brand new 3GHz iMac that’s had Time Machine running since it came out of the box. So no excuses that I can think of. Today I needed a file that got messed up since yesterday lunchtime, but there were no successful backups between 9am and 3pm, even though the Mac was running constantly and should have been backed up every hour. Time Machine is fantastic, but this needs fixing…
The article tells how to monitor Time Machine using the Console application. As an alternative to that method I’ve created a Unix alias that runs the following command in an xterm.
tail -F /private/var/log/system.log | grep backupd
This command does nothing special that Console doesn’t do. It’s just my preferred way on monitoring Time Machine.
The long thread about time machine backup failing on apple.com has been pulled. or someone got a link that work?