Video Screencast Help
Symantec to Separate Into Two Focused, Industry-Leading Technology Companies. Learn more.

B2D SATA / mount failures

Created: 26 Aug 2013 | 8 comments

Hi there,

I have BE2010 R3 and set up a few jobs, full and diff. The target is a UNC share that is always available. On that share I have created folders, for each job 1 folder. The target is a sata disk. The disk will be changed with another every monday. So full back-ups run in the weekend.

When I run the jobs manualy, they run fine. When leaving them scheduled, the fail.

Drive and media mount requested: 25-08-2013 01:00:00
V-79-57344-33039 - Error - Mount failed.
User canceled a Physical Volume Library operation.
 

The fulls and diffs have al the same media set, set to overwrite protection period 6 days and append period Infinite - Allow append.

The fulls and diffs are targeted to their own B2D on that drive, job set to Append to media, overwrite if no appendable media is available.

I did some searching and fount that the Append setting should be set to Overwrite.

I do not use a Device Pool since I have some enough articles that it also works like this setup.

So what is going wrong? For now I have them setup to Overwrite but do I need to set this up also with the Diff jobs? I don't wan't him to overwrite my diff every day. That should be kept for 1 week.

My goal is:

1. Run full jobs in the weekend and change the disk (UNC) every monday morning. It may ovewrite the old backup.

2. Run diff jobs every working day. Should overwrite after 1 week.
 

Is my setting wrong and if so, what should I change?

Thanks.

Operating Systems:

Comments 8 CommentsJump to latest comment

Jaydeep S's picture

Have you configured the Disk as a Removable B2D device?

On the B2D properties is the option "Allocate the maximum size for Backup-To-Disk files" enabled. If this is, disable it.

MarcelB's picture

I have not set it as Removable Device since I do not use a Device Pool. This option is there.

All Allocate maximum sizes are disabled (by default).

-

Marcel

pkh's picture

How is the disk attached to the media server?  When you specify the B2D folder, did you use the UNC path?

If you want to rotate your disks, you should try the procedure in my document below

https://www-secure.symantec.com/connect/articles/h...

I would suggest that you use 1 B2D folder per disk because if you do upgrade to BE 2012, it will only allow 1 B2D folder per disk.

MarcelB's picture

The disks are placed as an internal disk via the sata connector. We have a 5,25" tray build in where we can swap the disk easily. The disk has 1 folder that is shared. In this shared folder each job has its own directory. So yes a UNC path.

The article is great and indeed I used it. The only part not used is the device pool. So only configured with B2D folders per job, and 1 media set (not a pool).

I hope the Overwrite option is the one causing the problem since the rest run fine.

The only question is, do I have to set the diff jobs to Overwrite as well?

-

Marcel

pkh's picture

I know the device that you are talking about.  However, I don't think BE is designed to handle such devices.  It will not be able to detect that you have swapped the disk in the tray.  You would probably have to do an inventory after you have swapped the disk.

The overwrite option has nothing to do with your problem.  However, you should use overwrite in your jobs.  There is no advantage in appending to disk media.

Colin Weaver's picture

Not sure if this is the issue, but BE cannot handle B2D devices going offline and then online again  if they are NOT directly connected to the media server (but accessed over shares). I assume as you talk about configured via UNC that this Sata device is not directly connected. If this is the case then when changing drives you probably will need to restart the Backup Exec services in order to remain stable.

You would be better off if you can connect the device directly to the media server

Note: B2D devices configured on network shares should not be configured as Removable B2D anyway, even if the share is hedl on a removable devcie.

MarcelB's picture

OK. But the offline period I so short ... and how can we explain that a job running manualy runs fine and scheduled not? It could have something to do with the restart of services that is not happening?

-

Marcel

Colin Weaver's picture

If the drive is detected as offline at any point BE will not then notice if it comes back online (it would if it was directly connected but won't on a share) Whether or not BE decides a drive is offline depends slightly on what BE is doing at the time with regards other jobs and also partly on Device and Media maintaining a connection. As such how short the drive is offline may make no difference, just the timing of when it went offline might.

You can easily see if BE thinks the drive has gone offline by looking in the storage section (check maybe 5minutes after you disconnected the drive)

It should affect manual jobs in the same way if this is the problem (unless you restarted your services due to what happened overnight before you run the manual job)

Device and Media will not check for remote drives coming back online as we require an operating system volume insert notification to trigger that process and with the device being remote on a share, it is the remote operating system that gets this alert not the media server. Restarting the Device and Media Service (so all the BE services)  forces BE to check the status of all known devices which will bring it back online.