I needed to do a restore of some voicemails on a server and the calculations on how long it was going to take were slightly off as you can see from this screenshot - it actually got as far as 150 days before it reached the data on the file and started the restore off. Typically I have to do this when the voicemail server crashes (due to a power outage when it wasn't on a reliable UPS). Everytime the server crashes I would have to restore 2 peoples voicemails (out of about 70 voicemail boxes). They are the only two who have their voicemail wiped and Avaya don't care about the problem. What is weird is that this time the server had a controlled shutdown and not a sudden crash.
I sent the user this screenshot and told them it might take me a while but I'm working on it - they replied saying they didn't think they had *that* many voicemails!
Recently in Backup Category
I didn't know that you can only eject an ultrium tape drive when it is completely rewound. Detailed instructions from HP on removing an Ultrium drive are provided. Our Dell technician didn't know this and we were given the ability to remove the tape *any* way we liked when a tape was stuck in a drive that was replaced. Now I'd like to know if this would have solved the problem....
Thanks IT Technobabble
Fixed! One of my servers has been failing to backup with the error "0xe00084af The directory or file was not found, or could not be accessed. Final error category: Job Errors. For additional information refer to link V-79-57344-33967" I spent ages troubleshooting the errors and trying to work out what was going on and found that it would fail to backup any file on the local hard disk of the machine.
I posted a note in the symantec forums and didn't hear anything back, but did find a post that upgrading to 10d might fix it (not a current solution as this would mean purchasing an upgrade of the software for the exchange agent and the exchange agent is currently working)
The other solution was to stop SQL servers on the box. This server was the WSUS box and I had also recently upgraded it to version 3 of WSUS. This created (at least) two new services - SQL Server VSS Writer and Windows Internal Database (Microsoft ##SSEE). Through trial and error I discovered that stopping the SQL Server VSS Writer service meant the backup would work, which is weird as why this should affect me backing up something like c:\jobs\fred.bat which has nothing to do with SQLI don't know.
I'm hoping that my forum posting about the problem will get a better solution but for now I'm just pleased to be able to backup my file server.
One of our clients has been having issues with Volume Shadow Copy and SQL (MSDE) causing the backup to fail with the error 0x800423f4. By following the script at TazNetworks it looks like the backup might work tonight. It's too early to tell but at least when I kicked the job off briefly it started to some work rather than just quit out of ntbackup.
Update Unfortunately the solution didn't work for me, but I noticed that there was a hotfix available (904418) which is for x86 based operating systems. I noticed that they were trying to register the Sqlvdi.dll file, so I reran the registration of this dll (which wasn't in the original script above), restarted ntbackup and the system state backup worked ok which was great news as it wasn't working previously. I then kicked off the nightly backup to run though and it looks like it is working, so in short -
cd C:\Program Files\Microsoft SQL Server\80\COM
regsvr32 sqlvdi.dll
I've been trying to backup the main pc at home as it's getting really really slow for a 1 year old pc. Kristen blames it on all the stuff I keep installing on it which is probably true. The weird thing is that I can't see anything in task manager or perfmon that shows why it is so slow. CPU and memory usage is low yet the machine sometimes takes ages to respond so it's a reinstall time.
However, before doing that I wanted to image the hard disk - I've been attempting to do this since Saturday now. MicroCenter were doing a 160gb hard disk for $40 after rebates so I thought I'd ghost the original disk to my external usb disk and then restore the image to the second disk. (No link as the rebate was only valid until Saturday)
I had several attempts of doing this, but found that the pc would freeze whilst in the middle of the image. Yesterday morning I started the image off and when I came home from work it was still going.... until I shredded something and it crashed ghost with an error back to dos.
I then tried Acronis true image which said it would take 20 hours to ghost the image (of 40gb!) but again this morning it had frozen after running overnight.
By the end of last night I was really fed up with the thing as it shouldn't be this difficult so I went to sleep and decided to try again this evening.
I've a feeling that for some reason my external usb disk is being detected as a USB1 disk instead of USB2 which is why the original ghosting is taking forever.
I've now had the idea of ghosting the image direct to the new hard disk from the old one - if that works (which is giving me a realistic time of about an hour to ghost) I may be able to copy the ghost image back to the original disk and then restore it back over to the new disk.
Once I have a working ghost source I can scrub the new disk and reinstall from scratch and also see if I can get vista on the machine too......
We're back! My site was closed sometime last week as the server it was hosted on was having resource issues. I was offered a move to a new server and unfortunately the flat web pages got moved across but the sql database backend didn't. It's now been fixed and we're live again. Expect a couple of posts from me as I catch up on my blogging from the past couple of days.
Now I'm off to ensure I have a REALLY good sql backup of my data - for some reason my control panel backup didn't include a valid copy of my mysql databases.
Discovered today (after a lot of hassles) that Backup Exec may have a problem where backups to a disk folder does not work and ends up in a queued status. Eventually discovered that if you drag the "backup to disk" media to the scratch media pool then the backups start working again.
Didn't find anything on the veritas/symantec web site about this and nothing in google either so hopefully this will help other people out.
Kristen had an upsetting moment when she plugged in the XD card from the camera and saw the following
It looked like all the photos that she had taken today were corrupt and unreadable (and translated into chinese!). Fortunately I remembered reading about PC Inspector File recovery which I had tried out but not in a real life scenario. One download later and a search of the hard drive returned lots of image files stored on the disk (as expected). Recovering them was just a matter of selecting whereabouts I wanted to put the files (on a different disk). There were two options available. The first was using FAT1 and the second was to use FAT2. When I used FAT1 most of the files came back ok, but there were several files that came back slightly corrupted - this one is one of the worst
Some of the other photos have been uploaded to my Birdtable zoto gallery as Kristen took some excellent pictures of the birds on the birdtable today.
Had a customer whose ntbackups suddenly stopped working on Small Business 2003. On further investigation ntbackup was trying to write the backup to the miniqic type of tape drive yet the actual tape that was in the drive was the travan tape (bearing in mind these tapes have been in use for a long period of time). Apparently there is a known problem with NTBackup incorrectly choosing the backup tape type on a drive that supports multiple tape types (KB article 831664).
When a backup starts, it immediately fails with the error “The requested media failed to mount. The operation was aborted”, even if the tape is in the tape drive. Some tape drives support tapes with different block sizes. Windows Small Business Server 2003 backup chooses the first tape type available, which may not correspond to the tape type of the tape in the drive.
This problem is most commonly seen with the Travan 20 tape drive. Windows Small Business Server 2003 backup chooses the MiniQIC tape type instead of the Travan Tape type.
Applying the hotfix makes it work (and it didnt need a reboot). I'm mystified as to why its suddenly started doing this - I did apply another tape related hotfix about a week and a half ago but backups have run correctly since then. Very strange.
Lifehut has a good post on how to enable the automatic backup of Wordpress to an email system such as google. Worked great although I couldn't use the form of username+tag@gmail.com as the mail server that wordpress runs on does not allow + in the email address as strictly speaking its not rfc compliant. Yahoo groups also has this problem in that you can sign up for a mailing list with a + in the address but not use that address in their web interface.
STGViewer is a great utility for recovering individual files from the activesync backups that get created on your harddisk normally in the directory C:\Documents and Settings\username\Application Data\Microsoft\ActiveSync\Profiles\pdaname
Following on from my inital investigation into backing up a windows2000 server using ntbackup I have written a script that ignores the ntbackup catalog files and the internet explorer cache files. It uses reg.exe from the support tools. Reg.exe is native to xp, but part of the NT4 and Windows 2000resource kit. I used the version that comes with the windows 2000 resource kit.
Customer queried part of our backup routine asking about qmgr0.dat as ntbackup is unable to back the files up. Apparently qmgr*.dat files are part of BITS and therefore always in use. More importantly, the faq on mgr0.dat says that there is a registry key that you can use to tell ntbackup (and other backup software?) not to backup these files. This is a great answer to my problem of how to filter out valid skipped files not backed up when you backup c:\
Update You need to use regedt32 to create the registry entries and not regedit
The server that I had to do a full restore on yesterday was scary...After it was plugged back in the domain it would not allow clients to authenticate. I suspected that this was due to the fact that it was restored from a tape over 2 weeks old and therefore the SAM on this server was out of date with the SAM on the PDC. Trying to use netdom (the one from supplement 4 to the resource kit) kept on coming up with access denied errors and was therefore not much use. Overnight I scheduled a reboot of the pdc and did nothing else to the BDC as I was going to attempt a fix this morning...however when I got there, the users were working away - I guess the reboot had synched everything up properly (thankfully)
