Jump to content

Don Lee

Members
  • Content count

    564
  • Joined

  • Last visited

  • Days Won

    29

Everything posted by Don Lee

  1. I want to add that the UUID on some removable volumes has not been entirely stable for me. I cannot reproduce it now, but I had call to try to use the UUID to match up removable volumes, and I found that the UUIDs would sometimes change if the disk were unmounted and re-mounted. YMMV
  2. I want to emphasize that this bug may be quite serious. I did a "media set copy" and it only copied the most recent "backup"s. This is intermittent, as far as I can tell, but given the length of time o run a test, I am reluctant. To be clear, the original media set has 10 backups (sessions) and the resultant copied media set ends up with only one. The media set copy operation is not an every-day occurrence, but it should work reliably.
  3. ...incidentally, the current "new media" also has given me a hard time on disk media sets by stopping to wait for me to "specify" a new set via the file selection dialog.
  4. I would have wanted to do the "start new media set" option in a script, but the thing it does not allow me to do is control the name of the resulting media set. There is no "manual" way to do the "new media set", and no way to give it a specific name (number in brackets). I could use this option, but it would mean that I would have to toss my naming conventions on my backup sets, which I would rather not do. ("backup_name_yr [month]") The bug here seems serious. If I do a "copy media set" operation, I don't get all the data. I ahve not been able to reproduce this manually. I have only seen this in my automatic-monthly scripts. On education: Mentally, I use the term "snapshot" to refer to the data in the catalog that lists all the files that were present on the volume at the point in time that the backup was run. I use "backup" as the operation that creates the snapshot and snapshot data. It has no instantiation on the backup media. Session is the real confusion for me, because with Retro 6, it appeared to refer to a backup operation, but if so, would correspond to a snapshot, but it appears that it did not. The word appears to have mostly disappeared in Retro 8/9.
  5. I just ran an experiment with a copy script, and verified that if I have a set with multiple backups (snapshots), and only have some of them "retrieved", the copy operation will copy *all* of the backups (snapshots). This looks like a bug. Anyone else doing this, or seen this happen? Feel free to educate me on the difference between sessions, backups, and snapshots. I have never been clear on the terminology.
  6. What does this mean? extgTask: Source polling timed out after 30 mins snippet of log: These appear from time to time. ???
  7. I note that with backup scripts (not proactive) with more than one source, the logs and activity window do not list the client name for the sources. For instance, I back up the "/Users" directory on each of two clients and the server with a backup script "UserBack", and end up with 3 entries in the "Activities" window. snapshot attached. Note also that the logs do not list the client name, so it is hard to tell which "Users" is being backed up. -dgl-
  8. Please don't loose hope. Hope would be offended (I think). ;-> -dgl-
  9. Don Lee

    WebDAV Services

    I also wanted to use the WebDAV feature, and did some experimentation with Retro 9.0.x. What I found is that the OS of the machine running the Retro engine makes all the difference. If you are running Mac OS X 10.5.8, it doesn't work. The hangs and other problems don't happen every time, but are frequent enough that you can't rely on it. I also tried running the engine on Mac OS X 10.6, which seemed to work fine. From what I could tell, Retro relies entirely on the ability of the Mac OS X kernel to mount the WebDAV "share", and Retro simply treats it as just another file system. Up until 10.6, Mac OS X does not work very well. I did not try 10.7. Are you running 10.5, 10.6, or 10.7?
  10. A little more information about exactly what you are doing would allow those on this forum to make helpful comments. There is not enough information in your post to make any useful guesses as to what is happening. (i.e. engine (server) machine, speed, memory, media set residency, other things running on machine, network, etc) We know that retro is not *always* so slow, or it would be entirely unusable, and there are lots of users.
  11. The "Wakeon lan" feature has to be enabled on the machine. Check out the system prefs -> energy saver -> options -> "Wake for Ethernet network administrator access". I believe that a similar option exists on Windows.
  12. Don Lee

    16 threads?

    I have seen crashed of the engine when "too many" things are going on at once. More than one or two heavy activities also can reduce net performance. It is wonderful to be able to so a second operation when one is already busy, but the ability to do 10 or 20 at a time is of dubious value. I recommend limiting that to 2 or 3 threads.
  13. I just had one of my proactive scripts quit because it hit its "scheduled stop time". I wanted it to finish, so rather than create a new script, I thought I would add another schedule item to run it just for today, leaving the existing schedule and script intact. When I went to add another script entry for "10 AM to 3 PM/every day", I ended up with two schedule entries on the script, both saying "12AM-12AM/every day". Clearly the act of adding the additional script entry overwrote/corrupted the existing schedule entry. I have seen this before in other circumstances. In my experience, the scheduling machinery does not work well. A single entry generally works, but more than one is trouble. (Retro 9.0.2.107, Mac OS X 10.6.8)
  14. Two things: 1. As outlined above, you have some limitations on v6 (10.5 or earlier) and v9 only runs really well on 10.6 and better. You also have to run only one at a time. Running both on the same box is *possible*, but not particularly practical. I think you are better off keeping the old PPC machine around to run v6, and swapping the tape device, hoping that usage of the v6 retro is minimal. 2. When you rebuild the tapes, you get the "catalog" back, but in my experience, you lose all your snapshots. This means that you can search for files, but cannot restore to any one of the points in time on your backup set. This is sub-optimal, but I can live with it. I still believe that retro v9 is a big step forward. It would be nice to have it do a better job with v6 backup sets, but realistically, this need is more theoretical for me than actual. In fact I access the old tapes very, very rarely. The biggest issue is that if I *do* need them, the time it will take to retrieve the data is much longer than it shoud be, because the rebuild take so long.
  15. Media set copy will produce a writable backup set (media set), BUT the I believe that the metadata will not match between v6 and v9, so you will get the moral equivalent of a full backup on the copied set when you do your next incremental. I have done this with tape, not with dvd. Tape seems to work though you can read my whining on this forum about some of it. The mismatch of metadata is something that Retro can not really improve in my view, because it stores more metadata in v9, and v6 just doesn't. You can't go back and "fix" the v6 data, and you don't want to ignore the differences. Your best bet is to start new media sets in v9, and only use v6 sets if required.
  16. 9.0.2.107 I include my log (below). Note that there are quite a few "errors". The summary says "3 errors" on completion. (see snap - included) Which 3 are THE errors? How are they counted? -dgl- + Normal backup using Weekly_desktop at 6/4/12 (Activity Thread 2) To Backup Set v9_iCompute... - 6/4/12 9:58:21 AM: Copying Joy_HD_X on Joy 6/4/12 11:14:26 AM: Snapshot stored, 114.9 MB 6/4/12 11:14:34 AM: Comparing Joy_HD_X on Joy *File "Joy_HD_X/Library/Caches/com.apple.LaunchServices-023504.csstore": didn't compare *File "Joy_HD_X/private/var/log/system.log": different data size (set: 1641, vol: 1804) *File "Joy_HD_X/private/var/log/windowserver.log": different data size (set: 40808, vol: 40972) *File "Joy_HD_X/private/var/log/asl/2012.06.04.asl": different data size (set: 2230, vol: 2389) *File "Joy_HD_X/private/var/log/asl/StoreData": didn't compare *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Eudora Log": different data size (set: 406400, vol: 84601) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Eudora Settings": different resource size (set: 252001, vol: 252055) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Eudora Settings.bkup": different resource size (set: 251677, vol: 252037) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Eudora Statistics.xml": different data size (set: 12571, vol: 12573) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Personal Nicknames": didn't compare *File "Joy_HD_X/Users/cjl/Documents/eudora folder/UsageStats": different data size (set: 3080, vol: 4623) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Eudora Items/Plugins/Esp/Esp.log": different data size (set: 9996, vol: 10200) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/IMAP Folder/sysadmin1/Inbox/#&$*- Hidden": didn't compare *File "Joy_HD_X/Users/cjl/Documents/eudora folder/IMAP Folder/sysadmin1/Inbox/Inbox": didn't compare *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Mail Folder/In": different data size (set: 1437094, vol: 1425361) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Mail Folder/Out": different data size (set: 200079, vol: 198207) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Mail Folder/Trash": different resource size (set: 911495, vol: 913475) *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Spool Folder/In.temp": didn't compare *File "Joy_HD_X/Users/cjl/Documents/eudora folder/Spool Folder/Out.temp": didn't compare *File "Joy_HD_X/Users/cjl/Library/Caches/com.apple.Safari/Webpage Previews/114C2EFC50BDB16438695C35C14CF2E9.jpeg": different data size (set: 45005, vol: 45904) *File "Joy_HD_X/Users/cjl/Library/Caches/com.apple.Safari/Webpage Previews/114C2EFC50BDB16438695C35C14CF2E9.png": different data size (set: 132337, vol: 134638) *File "Joy_HD_X/Users/cjl/Library/Preferences/.GlobalPreferences.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.dock.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.iWork.Pages.plist": different data size (set: 9664, vol: 9330) *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.PreferenceSync.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.Preview.bookmarks.plist": different data size (set: 8923, vol: 8969) *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.Preview.plist": different data size (set: 6835, vol: 6625) *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.print.custompresets.plist": different data size (set: 15144, vol: 15148) *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.recentitems.plist": different data size (set: 24371, vol: 24557) *File "Joy_HD_X/Users/cjl/Library/Preferences/com.apple.Safari.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/com.nowsoftware.nowcontact.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/ByHost/com.apple.iCal.helper.001451173bcc.plist": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/FileMaker Preferences/FileMaker Pro 8.0A Prefs": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/Now Contactâ„¢ Prefs/Now Contactâ„¢ Prefs (4.0)": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/Now Up-to-Date Prefs/Now Up-to-Date Printing Prefs": didn't compare *File "Joy_HD_X/Users/cjl/Library/Preferences/Now Up-to-Date Prefs/QuickDay X Events": different data size (set: 5074, vol: 3511) *File "Joy_HD_X/Users/cjl/Library/PubSub/Database/Database.sqlite3": different data size (set: 76680192, vol: 76810240) *File "Joy_HD_X/Users/cjl/Library/PubSub/Feeds/1c35aca2530ae5680f0856e77db55d2feeb30bbd.xml": different data size (set: 25379, vol: 33798) *File "Joy_HD_X/Users/cjl/Library/PubSub/Feeds/600e4cdbdbd613b6f88c22fb55bf7b8d9dc3c251.xml": different data size (set: 43132, vol: 42224) *File "Joy_HD_X/Users/cjl/Library/PubSub/Feeds/73de2d277fc5a895287ab76a68c912be73a637b6.xml": different data size (set: 10875, vol: 11187) *File "Joy_HD_X/Users/cjl/Library/Safari/History.plist": different data size (set: 116897, vol: 113114) *File "Joy_HD_X/Users/cjl/Library/Safari/HistoryIndex.sk": different data size (set: 1548288, vol: 1499136) *File "Joy_HD_X/Users/cjl/Library/Safari/LastSession.plist": different data size (set: 597, vol: 899) *File "Joy_HD_X/Users/cjl/Library/Safari/WebpageIcons.db": didn't compare 6/4/12 11:56:38 AM: 3 execution errors Completed: 107197 files, 16.6 GB Performance: 326.4 MB/minute (272.1 copy, 407.5 compare) Duration: 01:58:17 (00:13:57 idle/loading/preparing)
  17. How about that. Indeed, there are exactly 3 "diffrernt resource size" errors above. So why is a "different resource size" an error, and "didn't compare" and "different data size" only warnings? Looks a little odd to me. Maybe there is a good explanation, though.
  18. This may be related to the "problem" I reported here. Firewire drives are "named" differently, because they are mounted differently by the OS. I have seen my firewire drives mounted with multiple names. When Retro picks up the name on creation of the media set it "fixes" it, and if on subsequent reboots the name changes, sometimes Retro can't find it. I ran some experiments on this, and got mixed results. It looks to me like Retro uses the posix path to ID the drive sometimes, and another mehchanisn (that is not path sensitive) other times. The Retrospect folks need to look into it. It will be a headache down the road.
  19. When I plug in my tape drive, the retrospect engine does not "see" it until I restart the engine. This may be as intended, but if so, is not mentioned in the documentation that I can find. The "rescan" operation did not appear to pick up the tape drive, but I didn't wait 2 minutes. ;-> Retro 9.0.1.401, Mac OS X 10.5.8, DDS USB tape drive. -dgl-
  20. I killed an activity today at 11:41 AM. The activity was part of a proactive script, that had 4 sources and one destination media set. I stopped the activity by pressing the "stop" button on the top of the activity window. This script, as I understand it, should have fired up the activity again, because it had only backed up about half of the volume at the time I stopped it. The Activity window lists this as being scheduled for 12:11 PM today (30 minutes after i killed it.) It has not run. Why? The activity window detail summary says under "status" "waiting to retry failed attempt". The script is active 24/7, and is not paused or otherwise disabled. The icon is the little "clock" with the green border. in the activity window. What is it waiting for?
  21. On restarting the engine, 3 of the activities started up. Something was "stuck" in the engine.....
  22. When I got up this morning, I checked, and _none_ of the proactive backup activities that should have happened over night happened. One backup that was scheduled for 4 AM was performed, but nothing else has been done. In fact, the last thing that was done (proactive) was that interrupted backup that is "waiting for retry". I'm going to restart the engine.....
  23. I got another crash of the engine last night. Retro 9.0.2.107 Mac OS X 10.6.8 Mac Mini 8GB mem from system.log: Jun 3 03:02:58 host-213 com.apple.launchd[1] (com.retrospect.retroengine[96947]): Job appears to have crashed: Segmentation fault Jun 3 03:02:59 host-213 com.retrospect.retroengine[82151]: #2> Command line is /Library/Application Support/Retrospect/RetrospectEngine.bundle/Contents/MacOS/RetroEngine Jun 3 03:02:59 host-213 com.apple.ReportCrash.Root[82150]: 2012-06-03 03:02:59.404 ReportCrash[82150:1407] Saved crash report for RetroEngine[96947] version ??? (???) to /Library/Logs/DiagnosticReports/RetroEngine_2012-06-03-030259_localhost.crash Jun 3 03:03:17 host-213 com.retrospect.retroengine[82151]: DAGServer:RegisterProtocal Protocal(GUID=200)! At the time of the crash, I had queued up several scripts that I had limited to "activity thread 3" so they would serialize. The scripts each did a recycle copy of a media set to another media set, with a recycle of the source at the end (on success). I was doing monthly rotations. (I still don't have a really *good* way to do these....) the crash log is attached.
  24. The garbage in the error message looks mighty familiar - much like what I see from my USB-connected DDS tape drive on errors.
×