tiffert Posted February 2, 2010 Report Share Posted February 2, 2010 (edited) Filevault not in use. All machines have an admin account. Toggling firewall off/on on both machines had no effect. The path to the missing XXXX volume was /Users/XXXX on XXX.local It is a user account folder (not an admin user). It is defined as a Favorite folder in Retrospect. Edited February 2, 2010 by Guest Quote Link to comment Share on other sites More sharing options...
CallMeDave Posted February 2, 2010 Report Share Posted February 2, 2010 It is defined as a Favorite folder in Retrospect OK, so that confirms R8's error reporting; it treats a Favorite Folder the way Retrospect Classic treated a Defined Subvolume; as a "Volume." I can SSH into the "missing" volume on the client machine, however, so I know it is on the network. Then at that point, can you go into the Retrospect Console and Browse the Favorite Folder? Can you Browse the Folder's enclosing logical volume? I mistakenly used "sleep" upthread when I should have used log out; but are you certain that the logged out machine isn't going to sleep via Energy Saver? Is this happening only on Scheduled Scripts? Or can you run a Script and observe it happening in real-time? Dave Quote Link to comment Share on other sites More sharing options...
tiffert Posted February 2, 2010 Report Share Posted February 2, 2010 (edited) The problem has suddenly resolved. Best as I can tell, the last thing I tried was to favorite the admin user account folder on the engine/console and client macs (even though this is not the folder I am trying to duplicate). If the problem recurs, I will report back. I regret that I don't have the time to systematically troubleshoot further at the moment. I am grateful for the feedback. Edited February 2, 2010 by Guest Quote Link to comment Share on other sites More sharing options...
Maser Posted February 3, 2010 Report Share Posted February 3, 2010 Huh... Now after never seeing this before... One of my client is currently logged out (no FileVault, no Firewall). 10.6.2. Running the 6.3.028 client -- *which I can refresh* from the sources list and get a different "speed" each time... And both my instances of Retrospect engine believe the client source is "unavailable" (meaning I can't browse the Macintosh HD") and the computer is not getting backed up.. And yet, I can control the machine (at the login screen) via ARD. and "top" says the "retroclient" process is running on that client: 61 retroclient 0.0 02:10.56 7 1 57 66 668K 288K 1644K 23M 601M 60 1 sleeping 0 2499004+ 4786 790686+ 390782+ 2408957+ 1109890+ 387905+ 122 root Weird... Quote Link to comment Share on other sites More sharing options...
rhwalker Posted February 3, 2010 Report Share Posted February 3, 2010 And what does the Retroclient status say when you open it in ARD? I guess you would have to log in to a local account, at least, to do this. Hopefully a local account that doesn't mount any volumes. Wouldn't it be nice if retroclient had a CLI interface for situations like this, so you could ssh in and check things in this exact status? Russ Quote Link to comment Share on other sites More sharing options...
Maser Posted February 3, 2010 Report Share Posted February 3, 2010 Interesting... Client said "Ready" when I logged into the machine (via the root account on the Mac -- over ARD.) I command-clicked "off" so the client was not running. Waited a couple of minutes. Clicked "On" -- client went to "Ready". AT THIS POINT -- I could browse the client "Macintosh HD" from the engine machine. (client said "connecting" and "in use..." as expected while I was doing this...) HOWEVER... At this point I logged off the root account. And I could no longer browse the HD (source unavailable). (Haven't rebooted the client yet -- which I know usually fixes the problem. I'm still trying to see if something obvious is going on looking through other log files...) Quote Link to comment Share on other sites More sharing options...
rhwalker Posted February 3, 2010 Report Share Posted February 3, 2010 And both my instances of Retrospect engine believe the client source is "unavailable" (meaning I can't browse the Macintosh HD") and the computer is not getting backed up.. I wonder if this is a case of competing Retrospect engines getting confused at this client. Your test, with multiple running Retrospect engines, might not be the same test as the original poster's. What happens if you stop one of the engines, repeat the experiment (ARD login, check status, logout) and then see if the client is still unavailable on the single running engine? Oh, and when the client is "unavailable", is port 497 open on the client? (check by telnet, and also perhaps by ssh in and netstat or some similar tool) Russ Quote Link to comment Share on other sites More sharing options...
Maser Posted February 3, 2010 Report Share Posted February 3, 2010 The second instance of the engine has no scripts. It's a copy of my other config80.dat file with only the sources that I use for testing. However, I tried it with one of the engines stopped -- same thing. Can browse if I'm logged in and turned the client off/on -- but can't browse once I log out. and, yes, 497 is open when it's at the login screen: Open TCP Port: 22 ssh Open TCP Port: 497 dantz Port Scan has completed… Maybe tech support was right and something about the recent 10.6.2 security update did something... Quote Link to comment Share on other sites More sharing options...
Moofo Posted March 24, 2010 Report Share Posted March 24, 2010 It may not be pertinent, but I have this exact same issue backing up serveral 10.6.2 server machines using the client with retrospect 7.6 for Windows. I'm leaving the console logged on for now, but I don't like it... Quote Link to comment Share on other sites More sharing options...
rdenis17 Posted August 31, 2010 Report Share Posted August 31, 2010 It may not be pertinent, but I have this exact same issue backing up serveral 10.6.2 server machines using the client with retrospect 7.6 for Windows. I'm leaving the console logged on for now, but I don't like it... I know that this is several months down the road. I'm wondering if anything new has been found on this issue. I have several Mac OSX 10.6.4 servers with client version 6.3.0.27 and newer that have to be left logged on with the client running on the server's console. I also have one 10.5.8 server that needs to be logged in for the scheduled backup to be successful. The only machines that seems to backup while logged out is a 10.4.11 server. Each site that has one of these OS X servers has its own Mac Mini running the Engine and Console. The servers are merely running the Retrospect Client. Each site has an appropriate license. When logged out on the server, the error message at the Retrospect Console is "Can't access volume "Volume name on Server Name", error -1101 (file/directory not found)". I am backing two complete volumes on each server. So, has any more been found out about this problem? Like Moofo, we are conducting the backups with a user logged in and the client running as an open application. But we would rather not. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.