Replaced dead external backup drive causing +1K errors

Server
1

I’m using v2.4.13 server on Win 10 Hm, with 5 clients v2.4.11 on mixed WinServer8 and WinServer12.
I was backing up to an external USB drive. The drive crashed, so I found and replaced it with nearly identical drive. Now I’m getting thousands of errors.

I’ve ran, multiple times, in admin acct: cleanup.bat, cleanup_database.bat, remove_unknown.bat, and repair_database.bat. This seemed to cleanup some errors, but still getting alot (see below).

I’m thinking about starting over from scratch, but is there way to start over from scratch without loosing server and client configuration?

On 4 clients: there are 3 errors:

Cannot get volume for path “”. The filename, directory name, or volume label syntax is incorrect. (code: 123)

Starting shadow copy for “.symlink_SqlServerWriter_{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}_35f247a261c6954158e3bd4167dcb5ac_master_files00000001” failed.

Error indexing files 1 of component “master” of writer “SqlServerWriter”.

And 1 client has over 32K errors:

Cannot copy directory metadata from “X:\UrBackup\office3\210509-2003.hashes.symlink_SqlServerWriter_{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}7230d34e3fffba69d9f76daab0b4cb38_master_files00000000.dir_metadata" to "X:\UrBackup\office3\210510-2010.hashes.symlink_SqlServerWriter{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}_7230d34e3fffba69d9f76daab0b4cb38_master_files00000000.dir_metadata”. - The system cannot find the file specified. (errorcode=2)

Cannot copy directory metadata from “X:\UrBackup\office3\210509-2003.hashes.symlink_SqlServerWriter_{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}7230d34e3fffba69d9f76daab0b4cb38_master_files00000001.dir_metadata" to "X:\UrBackup\office3\210510-2010.hashes.symlink_SqlServerWriter{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}_7230d34e3fffba69d9f76daab0b4cb38_master_files00000001.dir_metadata”. - The system cannot find the file specified. (errorcode=2)

  • 30K similar errorsdir
2

Are these for full backups or incrementals?

Have you done new full backups since the backup drive was replaced?

3

Good thought. The Client still knows it has sent previous backups, so forcing a new Full Backup from the Server may be necessary. If doing this manually (from the Status page ▼ drop-down menu) is impractical, I suggest turning off Incremental Backups in Settings: File Backups and check Disable after the Interval for incremental backups. Temporarily set the Interval for full file backups to 1 days to ensure UrBackup tries as soon as possible.

Repeat the above for Image Backups if those are also failing.

2 Likes
4

@BrainWaveCC
Yes, I individually started full backup on each client, and of course errors as well, then each night does the incrementals…

@Don_Wright
I did the full backup from the Status page. I will try as you suggested; temporarily disable incremental, and let full file backups to autostart.

Thanks.

5

I’ve turned off Incremental File Backups, set Full File Backup Interval to 1 day, and increased the min and max Full File Backups. Waited for it to start the Full File Backup processes for the clients when time came (8PM every night), but it never started, and no log to indicate that it attempted to backup. So on 2nd day, I restarted the server, but still no go. After couple of days, I went ahead and started the Full File Backups manually, still Incremental disabled, but all failed w/ following errors:

backupcom->StartSnapshotSet(&dir->ref->ssetid) failed. VSS error code VSS_E_SNAPSHOT_SET_IN_PROGRESS

Cannot get volume for path “”. The filename, directory name, or volume label syntax is incorrect. (code: 123)
Starting shadow copy for “.symlink_SqlServerWriter_{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}_35f247a261c6954158e3bd4167dcb5ac_master_files00000001” failed.
Error indexing files 1 of component “master” of writer “SqlServerWriter”.

Not sure where to go from here. Advice would be appreciated.
Meanwhile, I’m gonna replace the hard drive to see if that makes a difference.

6

What is different between the old drive and the new one?

I’d recommend that you try the following on the server:

https://www.urbackup.org/administration_manual.html#x1-750008.6

The remove-unknown.bat script may be helpful.

7

After running cleanup.bat, cleanup_database.bat, remove_unknown.bat, and repair_database.bat, I ran the Full File Backup manually once more, and to my surprise, many of the errors has gone away!! Few clients completed Full File Backup without any error/warning!! Couple clients still had few errors and warnings, but some of those errors cleared out after subsequent Incremental File Backups.

The below are the remaining errors and warnings on couple of clients. At least, I’m backing up now, just need to clear some of these errors/warnings…

Errors 05/18/21 02:02 Cannot get volume for path “”. The filename, directory name, or volume label syntax is incorrect. (code: 123)
Errors 05/18/21 02:02 Starting shadow copy for “.symlink_SqlServerWriter_{A65FAA63-5EA8-4EBC-9DBD-A0C4DB26912A}_35f247a261c6954158e3bd4167dcb5ac_master_files00000001” failed.
Errors 05/18/21 02:02 Error indexing files 1 of component “master” of writer “SqlServerWriter”.
Warnings 05/18/21 02:16 Error adding sparse file entry. Could not read metadata from X:\UrBackup\client1\210518-0141.hashes\Data\Share\cat\1533567927057.jpeg
Warnings 05/18/21 02:16 Error adding sparse file entry. Could not read metadata from X:\UrBackup\client1\210518-0141.hashes\Data\Share\cat\1533733753717.jpeg

… continues 190 more…

Warnings 05/17/21 20:08 Writer NTDS has failure state VSS_WS_FAILED_AT_POST_SNAPSHOT with error VSS_E_WRITERERROR_NONRETRYABLE. UrBackup will continue with the backup but the associated data may not be consistent.
Warnings 05/17/21 20:08 Writer is in error state during snapshot creation. Writer data may not be consistent. This means the files open by this application (e.g. databases) will be backed up in a crash consistent state instead of a properly shutdown state. Properly written applications can recover from system crashes or power failures.
8

Awesome!

There was some database vs file system inconsistency, and the clean-up scripts addressed it.

Some of those VSS errors may be transient. See when the next backup time periods come around.

-ASB

1 Like
9

Your were right about the VSS errors, they eventually went away. Now I only got 1 error, which seems to be consistent everyday. Is the error below related to possibly datafile left open at one of the workstation? It’s not uncommon for one of the lady’s to forget to close the application at the end of day. Thanks.

Errors 05/23/21 02:33 Error getting file patch for “TLCTiMK95hJXUmEsWhrG|eProp/eprop.db” from client3. Errorcode: TIMEOUT (2)

Errors 05/23/21 02:33 Client client3 went offline.

Errors 05/23/21 02:57 Backup failed

10

Very glad to hear that the scripts addressed your issues.

This error looks more like the machine was turned off (or went to sleep) before the backup completed. The only thing I find that’s weird is that the final “backup failed” message doesn’t occur until ~24 minutes after that “offline” message.

Are the other failures happening at the same time of day?

1 Like
11

UrBackup is persistent and waits for the connection to resume before declaring the backup failed, in case it was a Windows reboot for updates or similar. It will also try to resume a failed-for-timeout backup after the Client next connects to the Server.

2 Likes
12

So far, that same error appears twice every single night at various times, only for that one client. The client’s PC is a Win Server 2016 Standard, and stays on 24/7. Looking client’s side logs, it has 3 logs per night, 1st two failed errors, but 3rd one succeeded. It seems the backup fails then succeeds on 3rd attempt. At my end, I was looking at the Warnings and Error logs, and failed to look at the Info log which showed eventual succeeded backup. Problem solved!! Thank you all for your help! Greatly appreciated.

1 Like
13

It might be a Windows power saving issue. I support a client/server business program that claims we must turn off Power Management on the network interface card or various bad things will happen.

Open Device Manager (not Network and Internet) and check the settings for your network interface. If you find something like “Allow Windows to turn off this device to save energy” then disable that feature and see if things improve. Do this on both the Server and the Client. You might go so far as to turn off most power saving features using the High Performance power plan (which disables Sleep) and the available advanced settings that come with it.

1 Like
14

Good idea.

@itchibahn, it might also be a power saving issue on the network card itself. I always turn those off.

That would be in the NIC settings, if the place where @Don_Wright suggested doesn’t turn anything up.

-ASB

1 Like