Looks like no one’s replied in a while. To start the conversation again, simply ask a new question.

Time Machine Not Backing Up New Files

Time Machine is NOT backing up my recent files.


I recently noticed that Time Machine was backing up really quickly instead of taking several minutes as usual. I checked previous backups and discovered that NONE of my recent files were backed up.


I have a very basic setup. An external usb 1 tb drive formatted as Mac OS Extended (Journaled) being used as the time machine backup drive. A new i5 mac mini with OS X 10.8.4. About 200 gigs of data with one user on the internal drive.


I have followed all of the Pondini troubleshooting steps. I have restarted into safe mode to force a deep traversal. I have repaired permissions and verified all disks with Disk Utility. I have deleted time machine preferences and followed instructions as per Pondini #A4. The backup seems to run but then finishes quickly but DOES NOT BACK UP any recent files.


I have run TMUTIL COMPARE, it showed that many files needed to be backed up.


I have run TMUTIL ISEXCLUDED on some of the files that were not being backed up. It showed that they were not flagged for exclusion.


I have run a manual backup after starting up in Safe Mode, which forced a deep event scan, but then exhibited the same behaviour..


I have acquired a new external hard drive, backed up the complete main drive (using Time Machine which appeared to be successful on the first time full backup), then run a manual backup on that new drive after adding new files and had THE SAME PROBLEM.


Here's the latest console output, which is typical since I have noticed this issue. There appear to be many megabytes of new files to be backed up, but then the operation completes within seconds and no files are backed up at all.


6/23/13 7:15:16.422 AM com.apple.backupd[643]: Starting manual backup

6/23/13 7:15:16.966 AM com.apple.backupd[643]: Backing up to: /Volumes/Omnius/Backups.backupdb

6/23/13 7:15:17.916 AM com.apple.backupd[643]: Using file event preflight for Olympus

6/23/13 7:15:20.277 AM com.apple.backupd[643]: Will copy (375.1 MB) from Olympus

6/23/13 7:15:20.298 AM com.apple.backupd[643]: Found 845 files (375.1 MB) needing backup

6/23/13 7:15:20.311 AM com.apple.backupd[643]: 1.91 GB required (including padding), 747.18 GB available

6/23/13 7:15:29.264 AM com.apple.backupd[643]: Copied 400 files (1.4 MB) from volume Olympus.

6/23/13 7:15:29.298 AM com.apple.backupd[643]: Using file event preflight for Olympus

6/23/13 7:15:29.299 AM com.apple.backupd[643]: Will copy (Zero KB) from Olympus

6/23/13 7:15:29.300 AM com.apple.backupd[643]: Found 5 files (Zero KB) needing backup

6/23/13 7:15:29.301 AM com.apple.backupd[643]: 1.46 GB required (including padding), 747.17 GB available

6/23/13 7:15:30.439 AM com.apple.backupd[643]: Copied 140 files (33 bytes) from volume Olympus.

6/23/13 7:15:30.511 AM com.apple.backupd[643]: Created new backup: 2013-06-23-071530

6/23/13 7:15:30.608 AM com.apple.backupd[643]: Starting post-backup thinning

6/23/13 7:15:30.608 AM com.apple.backupd[643]: No post-back up thinning needed: no expired backups exist

6/23/13 7:15:30.689 AM com.apple.backupd[643]: Backup completed successfully.


Please help. Time Machine is currently useless as a backup solution.

Mac mini (Mid 2011), OS X Mountain Lion (10.8.4)

Posted on Jun 23, 2013 7:31 AM

Reply
104 replies

Jul 3, 2013 1:43 PM in response to O157-H7

O157-H7 wrote:

I've already tried swapping in an empty TM target drive. That worked, but then it went back to not backing up some new files and changes.

Then those rsync scripts are likely the cause. Rsync is not a "few lines of perl". It is 3.2 MB of C source. It is wrtten by GNU and they are not concerned with either Mac or Time Machine compatibility. What it sounds like is that the OS detects changes in those directories but when it goes to check, discovers that their timestamps are older than those in Time Machine, so it ignores them. Once that happens, they aren't ever going to get backed up until you blow away your entire Time Machine drive. If my theory is correct, then deleting those .fseventsd file would not be sufficient to correct the problem.


I would not recommend running rsync with Time Machine. I suggest you find a new way to perform your network backups. A few lines of Perl is probably a good idea, as long as rsync isn't involved. You can keep searching for an answer that allows you to blame Mountain Lion if you want. I'm satisfied I have solved your problem, but only SeattleMacUser can definitively say for sure.

Jul 3, 2013 2:30 PM in response to etresoft

Then those rsync scripts are likely the cause. Rsync is not a "few lines of perl". It is 3.2 MB of C source. It is wrtten by GNU and they are not concerned with either Mac or Time Machine compatibility. What it sounds like is that the OS detects changes in those directories but when it goes to check, discovers that their timestamps are older than those in Time Machine, so it ignores them. Once that happens, they aren't ever going to get backed up until you blow away your entire Time Machine drive. If my theory is correct, then deleting those .fseventsd file would not be sufficient to correct the problem.


I would not recommend running rsync with Time Machine. I suggest you find a new way to perform your network backups. A few lines of Perl is probably a good idea, as long as rsync isn't involved. You can keep searching for an answer that allows you to blame Mountain Lion if you want. I'm satisfied I have solved your problem, but only SeattleMacUser can definitively say for sure.


You'd be wrong. Look, I realize that many of the people that come here looking for help are not really familiar with CLIs or the complex underpinnings of OS X. As should be clear from my posts in this thread, I am not that kind of user. The timestamp issue was literally the first thing that ocurred to me, so I modified both the file creation and modification dates, and they are not the responsible for the problem.


You also misunderstand my statement. My backup scripts are a few lines of perl that call rsync, which send any changes on the shared drive to an offsite server. But this is besides the point. The offsite rsyncs are not the issue. As I said in the previous post, TM worked fine before. There is some change in the time machine implementation in 10.8, or some other change in the filesystem that is causing the problem.


Please, read carefully what I have already posted in the thread.


When I get back, I'll in all likelyhood try and resurrect the old 10.5 machine and do a side-by-side comparison. Hopefully that will provide some clues as to what's going on.

Jul 3, 2013 5:26 PM in response to etresoft

G'day


1. I had in mind to develop the utility that I spoke of for checking that files are backed in TM and put it in the App store.But from Lion on Java isn't installed on Macs by default so lots of people won't now have it. I'm not about to learn Objective C just for one small utility.


So there is less value for me to make it. If you are serious about making it then maybe these thoughts would help. I was thinking about 3 threads (possibly literally) in the utility. The first would: run continuously; sleep till it finds a new TM backup; get the list of recently changed files; divide them into 60 lumps and process one lump a minute (to limit system impact), ensuring that each file was in the latest TM backup; any file that isn't would have to be checked to ensure it hadn't been deleted, isn't newer than the backup, is otherwise within the scope of TM etc before being brought to the users attention; in bringing to the users attention you'd maybe want to filter the list - no point in listing 1000 files from the same directory. The second thread would: iterate from 0 upwards; get the list of files modified in that many days prior to today; remove from the list any files already processed; then do as thread 1; at some point it would have to set the last modified date to infinity to get all files so it can eventually terminate; this thread would run once and probably never run again execpt on request. The third thread would do as the second but would only look at files if the (hash of) the list of files modified (on or since) day n differs from that found by thread 2 or as updated by thread 3 earlier; on exhausting the list of hashes thread 2 would start over.


The purpose of thread 1 is to give early warning that recently changed files haven't been backed up by TM. The purpose of thread 2 is to do a full system sweep looking for files not in the backup. The purpose of thread 3 is to check files that are added to the system with a back-dated last modified date, as happens with software installs, untars etc. One obvious gotcha - thread 3 has to make sure it compares against the right hash from thread 2 since n days ago is a relative term. Thread 2 will also need to replace hashes it found to differ.


2. I thought you were a bit hard on O157. I found his comments relevant to the discussion and helpful.


3. For O157. Its still possible that the rsync is the problem. That you still have the TM problem after stopping rsyncs isn't diagnostic. If rysncs run in the past caused some kind of folder damage it may be that that damage, which persists though you have stopped running rsyncs, is causing the TM problem. The fact that newly added files are not being backed up may simply mean they were added to folders already damaged by rsync. Folder damage certainly looks like the or a trigger for the TM problem both from what I am seeing and from Pondinis contributions on how to fix the TM problem. (One thing you might do is: find a file that was recently added that TM didn't backup; put another new file in the same folder; if TM doesn't back that up then folder damage looks increasingly likely as a cause; if TM does back it up its something else.)

Jul 3, 2013 7:05 PM in response to Ian Blavins

G'day


Well there is good news and bad news.


I recreated my desktop folder. I was making the operation more complex than it needed to be in my thinking above. In the end I: did a full (non TM) backup; created a folder /Users/Ian/D2; in terminal moved each folder and file individually from /Users/Ian/Desktop to D2 (there were only about 15, being a tidy person); logged out of the Mac; logged into the Mac as Fred; in Fred logged into terminal as Ian; deleted /Users/Ian/Desktop with sudo; logged into the Mac as Ian, thereby creating a new Destop; logged into terminal as Ian; individually moved each file and folder from D2 back to the new Desktop; logged out as Ian; logged back into the Mac as Ian; and all seems to be well in terms of using the Mac.


I then put a new folder, lets call it B, into A (the folder on the Desktop whose new contents were not being backed up by TM) and a new folder C on the desktop. I then did a TM backup. Neither B nor C was backed up. So no change to behaviour. So replacing the Desktop folder hasn't fixed anything.


But in my work in terminal I spotted a file called .com.apple.timemachine.supported. So I put a copy of that in /Users/Ian/Desktop and did another TM backup. No change.


Then I put a copy of .com.apple.timemachine.supported into /Users/Ian and did another backup. This time B and C were backed up. So maybe a missing .com.apple.timemachine.supported file in /Users/Ian was the cause of the problem of files in that directory and below not being considered by TM as needing backup. I guess this is 'folder damage' in a sense.


Just to be sure I then put a new folder called D on the desktop and removed the above com file from /Users/Ian and did another TM backup. I was expecting D not to be backed up. But it was. I did a bit more experimenting with com files in /Users, /Users/Ian, and /Users/Ian/Desktop. It begins to look at though, once a directory has had a com file in it TM remembers that fact and the com file doesn't actually need to be there any more. Or maybe something else I did while testing caused TM to backup folders it was previously ignoring.


So I think I am warm, but not quite there yet. I will think about this some more, let thing settle, and dream up some more experiments.

Jul 3, 2013 8:10 PM in response to Ian Blavins

This discovery intrigued me, so I thought I'd give it a try via ssh. I've seen those hidden files from time to time, but haven't paid them much mind. I'm curious who is the owning user of your .supported files, and what are the permissions on the files?


Just to test, I touched a .com.apple.timemachine.supported file in a directory containing stuck items, and for good measure in every enclosing directory above it. (e.g. a .com.apple.timemachine.supported in /Volumes/LabDrive/users/username/Projects/escrt/testing, /Volumes/LabDrive/users/username/Projects/escrt, /Volumes/LabDrive/users/username/Projects, etc. etc.)


I triggered a manual TM backup after the creation of each .supported file, but to no avail. Oh well. I too will continue to ruminate on potential experiments.

Jul 6, 2013 11:51 AM in response to Ian Blavins

Hi everybody, I'm back. Sorry for the delay in following up. I was kind of hoping one of you other thread posters would shed some more light on the situation, but oh well.


So I did a complete image restore as suggested by Linc Davis. Here's how it went down:


- Backed up my drive to a secondary external drive just for safety's sake

- 'Touch'ed all files to force a complete Time Machine backup (as previously discussed in this thread)

- Ran Time Machine manual backup

- Restarted in Recovery Mode

- Used Disk Utility to erase and partition my internal startup drive

- Restored from Time Machine Backup

- Created new files & placed them in various locations in Home folder

- Ran Time Machine manual backup again (this took a LONG time, just over an hour. I've included the console log below)

- Checked newly created files, they WERE successfully backed up.


I will follow up with another report after I've changed and created files over the course of a day and run Time Machine again, just to be sure this wasn't a first-time occurrence only. I hope this will help you guys figure out what's going on here. As another poster pointed out, this issue subverts the entire purpose of Time Machine without giving any indication that it is failing. Not good.


7/6/13 10:37:56.169 AM com.apple.backupd[383]: Starting manual backup

7/6/13 10:37:56.741 AM com.apple.backupd[383]: Backing up to: /Volumes/Omnius/Backups.backupdb

7/6/13 10:37:57.775 AM com.apple.backupd[383]: Inherited root volume Olympus, UUID: F4A5C913-724C-3F89-B88E-13020463E232

7/6/13 10:37:57.873 AM com.apple.backupd[383]: Event store UUIDs don't match for volume: Olympus

7/6/13 10:37:57.946 AM com.apple.backupd[383]: Deep event scan at path:/ reason:must scan subdirs|new event db|

7/6/13 10:37:57.947 AM com.apple.backupd[383]: First backup after disk inheritance for / - complete scan required

7/6/13 10:54:18.413 AM com.apple.backupd[383]: Finished scan

7/6/13 10:54:27.415 AM com.apple.backupd[383]: Found 144 files (43.2 MB) needing backup

7/6/13 10:54:27.550 AM com.apple.backupd[383]: 1.29 GB required (including padding), 642.5 GB available

7/6/13 11:37:59.737 AM com.apple.backupd[383]: Copied 53.7 MB of 53.7 MB, 1122250 of 1122250 items

7/6/13 11:38:14.057 AM com.apple.backupd[383]: Copied 1126331 files (53.7 MB) from volume Olympus.

7/6/13 11:38:14.252 AM com.apple.backupd[383]: Using file event preflight for Olympus

7/6/13 11:38:15.726 AM com.apple.backupd[383]: Will copy (1.4 MB) from Olympus

7/6/13 11:38:15.730 AM com.apple.backupd[383]: Found 38 files (1.4 MB) needing backup

7/6/13 11:38:15.731 AM com.apple.backupd[383]: 1.37 GB required (including padding), 641.5 GB available

7/6/13 11:38:29.738 AM com.apple.backupd[383]: Copied 305 files (1.4 MB) from volume Olympus.

7/6/13 11:38:30.847 AM com.apple.backupd[383]: Created new backup: 2013-07-06-113829

7/6/13 11:38:32.111 AM com.apple.backupd[383]: Starting post-backup thinning

7/6/13 11:38:32.111 AM com.apple.backupd[383]: No post-back up thinning needed: no expired backups exist

7/6/13 11:38:32.490 AM com.apple.backupd[383]: Backup completed successfully.

Jul 7, 2013 2:54 PM in response to Linc Davis

/dev/disk0

#: TYPE NAME SIZE IDENTIFIER

0: GUID_partition_scheme *500.1 GB disk0

1: EFI 209.7 MB disk0s1

2: Apple_HFS Olympus 499.2 GB disk0s2

3: Apple_Boot Recovery HD 650.0 MB disk0s3

/dev/disk1

#: TYPE NAME SIZE IDENTIFIER

0: Apple_partition_scheme *1.0 TB disk1

1: Apple_partition_map 32.3 KB disk1s1

2: Apple_Driver43 28.7 KB disk1s2

3: Apple_Driver43 28.7 KB disk1s3

4: Apple_Driver_ATA 28.7 KB disk1s4

5: Apple_Driver_ATA 28.7 KB disk1s5

6: Apple_FWDriver 262.1 KB disk1s6

7: Apple_Driver_IOKit 262.1 KB disk1s7

8: Apple_Patches 262.1 KB disk1s8

9: Apple_HFS Omnius 1.0 TB disk1s10

/dev/disk2

#: TYPE NAME SIZE IDENTIFIER

0: GUID_partition_scheme *750.1 GB disk2

1: EFI 209.7 MB disk2s1

2: Apple_HFS Erasmus 749.8 GB disk2s2



No CoreStorage logical volume groups found

Jul 7, 2013 3:16 PM in response to Linc Davis

Will do. It is quite possible that this drive was acquired at the same time that I started having this problem.


I think that was a "Mac-Ready" Western Digital USB drive. I'm sure I erased it with Disk Utility upon purchase to get rid of the consumer cruft, but I am also pretty sure that I did NOT repartition it at the same time.


I'll erase AND repartition the backup drive, then run a full manual backup via Time Machine again this evening.


Then I'll try out the usual tests (recent files etc.) & let you know how it goes.

Jul 8, 2013 4:00 PM in response to SeattleMacUser

OK, after erasing AND repartitioning the backup USB drive, I ran a Time Machine manual backup which archived the entire startup drive anew to the backup drive.


(Just to be clear about this part, when I previously rebuilt the Time Machine archive from scratch - onto the drive with the older 'screwed up' partition table structure - it did NOT fix the issue.)


I used my computer normally today, creating and altering many files throughout my Home folder.


I ran a manual Time Machine backup again and saw the progress bar behaving normally (i.e. reporting several hundred megabytes of new files, then progressing slowly from start to finish as it backed them up) instead of skipping quickly from the start to the end after just a few hundred k as it had been doing previously.


I checked all of the recently created or updated files, and they WERE all successfully backed up.


It appears that correcting the partition table on the backup drive may have solved the problem. 😀


I will run a few more backups over the next 24 hours and make sure things continue to be backed up correctly.


Here is the current result of the 'diskutil list' command for your reference:


/dev/disk0

#: TYPE NAME SIZE IDENTIFIER

0: GUID_partition_scheme *500.1 GB disk0

1: EFI 209.7 MB disk0s1

2: Apple_HFS Olympus 499.2 GB disk0s2

3: Apple_Boot Recovery HD 650.0 MB disk0s3

/dev/disk1

#: TYPE NAME SIZE IDENTIFIER

0: GUID_partition_scheme *1.0 TB disk1

1: EFI 209.7 MB disk1s1

2: Apple_HFS Omnius 999.8 GB disk1s2

/dev/disk2

#: TYPE NAME SIZE IDENTIFIER

0: GUID_partition_scheme *750.1 GB disk2

1: EFI 209.7 MB disk2s1

2: Apple_HFS Erasmus 749.8 GB disk2s2



No CoreStorage logical volume groups found

Time Machine Not Backing Up New Files

Welcome to Apple Support Community
A forum where Apple customers help each other with their products. Get started with your Apple ID.