DailyBackups Extremely Inconsistent, Sometimes a Month Apart

[Modify_inc]

Hello!

I'm using Retrospect OEM Single Server v8.5.0(136) on Windows Server 2012 R2

Usually we just leave Retrospect to do its thing every day and have had no real issues. We recycle between two Dell R7X91 320GB cartridges for our daily backups named, DailyBackup1 and DailyBackup2 starting at 8pm M-F. I always take one home at the end of the day, then swap it out for the previous night's backup when I return in the morning.

I just happen to check the history today and noticed there have been many, I mean many days that show Execution Incomplete. Also, no data shows to have been copied over. Completed 0 files, @ zero KB, Remaining 298711 files @ 278.3 GB. The errors below are fairly the same for all of them, except there are a few where some data is copied, yet it still says the media request timed out after waiting, which makes no sense to me. How can anything be copied if the error just before stated the media request timed out?

So every day (M-F) there are two logs for every work day. The first is usually a timeout error because the cartridge or DailyBackup it expects is not available (it's offsite), so then it looks for the other DailyBackup. It finds it and then starts backing up. Well that is what is used to do. Now, it just seems terribly inconsistent and I have no idea why or what has changed.

The failed backup errors below are from 8/3/17. It wasn't until 8/18/17 (2 weeks later) that the log showed a full backup.

Nothing Backed Up

+    Recycle backup using PP_Backup at 8/3/2017 8:00 PM (Execution unit 1)
    To Backup Set DailyBackup1...

-    8/3/2017 8:00:53 PM: Copying DATA (P:)
        Using Instant Scan
        Media request for "2-DailyBackup1" timed out after waiting 0:01:00
    8/3/2017 8:02:42 PM: Execution incomplete
        Remaining: 298711 files, 278.3 GB
        Completed: 0 files, zero KB
        Performance: 0.0 MB/minute
        Duration: 00:01:49 (00:01:14 idle/loading/preparing)
 

+    Recycle backup using PP_Backup at 8/3/2017 8:02 PM (Execution unit 2)
    To Backup Set DailyBackup2...  -    8/3/2017 8:02:43 PM: Copying DATA (P:)
        Using Instant Scan
        Media request for "2-DailyBackup2" timed out after waiting 0:01:00
    8/3/2017 8:04:28 PM: Execution incomplete
        Remaining: 298711 files, 278.3 GB
        Completed: 0 files, zero KB
        Performance: 0.0 MB/minute
        Duration: 00:01:45 (00:01:11 idle/loading/preparing)
 

Full Backup

+    Recycle backup using PP_Backup at 8/18/2017 8:00 PM (Execution unit 1)
    To Backup Set DailyBackup1...

-    8/18/2017 8:00:54 PM: Copying DATA (P:)
        Using Instant Scan
    8/18/2017 11:35:12 PM: Snapshot stored, 52.6 MB
    8/18/2017 11:35:16 PM: Comparing DATA (P:)
    8/19/2017 2:52:33 AM: 65 execution errors
        Completed: 302090 files, 280.2 GB
        Performance: 1402.3 MB/minute (1348.5 copy, 1460.7 compare)
        Duration: 06:51:38 (00:02:34 idle/loading/preparing)

Partial Backup

+    Recycle backup using PP_Backup at 8/22/2017 8:00 PM (Execution unit 1)
    To Backup Set DailyBackup1...
   
-    8/22/2017 8:00:55 PM: Copying DATA (P:)
        Using Instant Scan
        Media request for "2-DailyBackup1" timed out after waiting 0:01:00
    8/22/2017 8:16:45 PM: Execution incomplete
        Remaining: 243204 files, 271.1 GB
        Completed: 59602 files, 17.3 GB
        Performance: 1220.1 MB/minute
        Duration: 00:15:50 (00:01:19 idle/loading/preparing)

So what could cause this, where or what do I need to check? I'm also certain a cartridge is in the drive at all times as I am the one doing it. I'm still really puzzled why a few of the logs indicate a transfer started but then only a fraction of the files were actually copied, leaving 90% or more of disk space.

operations_log.utx

[DavidHertzberg]

Modify_inc,

A fair amount of Googling for "Dell R7X91" makes it abundantly clear that this cartridge is a Dell name for some model of RDX cartridge—which Dell used to call RD1000.  Since that is so, I strongly recommend that you read this post and make sure that you have not defined your Backup Sets as Removable Disk or File.  If you have, redefine the Backup Sets as Disk, and then read the rest of the posts in that thread for further enlightenment while you are rerunning your backups.

P.S.: Nice OP, BTW; you supplied nearly all the necessary information—which many people making an OP on these Forums don't do.  I now see that in this 2015 post you say  (presumably Retrospect Technical) "Support" told you to define your Backup Sets as "fixed"—whatever that means (it's not a Retrospect Windows choice AFAIK).

Edited January 25, 2018 by DavidHertzberg
Defining an RDX cartridge as a File Backup Set isn't a good idea either; Dell used to call its RDX cartridges RD1000; P.S.

[DavidHertzberg]

Modify_inc,

Another possibility is that your RD1000 dock has gotten old, so that the connector to your cartridges is now occasionally flakey.  Maybe you're not always pushing the cartridge solidly in to the dock.  Or maybe the connector on one/both of the cartridges is flakey.  Just guessing from what I've read of RDX , since I've never owned an RDX/RD1000 dock-cartridge system.

[mbennett]

I'm seeing a lot of errors showing "can't read error -1017 insufficient permissions" in the log file you attached.  First thing is to reboot the system that's being backed up and run chkdsk /f on that drive.  That might fix the problem.  The insufficient permissions error probably refers to the directory where the catalog files are located, so you should also try rebuilding all of the catalog files that are concerned in the jobs that aren't working.

Also, look up that -1017 error in the knowledgebase on the Retrospect website and see what it says. I didn't do that.

It's always possible that the RDX drive is worn out or a cable is bad, but they're pretty reliable. I use one every day. It's also possible the RDX cartridge itself is bad.

Do you have any grooming jobs running on these sets?