Overenthusiastic connections to missing NAS File media set

[prl]

I'm transitioning from doing backups on optical media to doing backups on file media sets on a NAS.

 

Yesterday, my NAS server died (it's a server on my ADSL modem/router, which succumbed to a voltage spike in a thunderstorm).

 

Now, when I try to do a backup on the optical media sets (not using the NAS backup sets at all), Retrospect keeps trying repeatedly to connect to the NAS that I don't want to use. Each time it tries to connect, I have to dismiss a warning box 12 times (once plus five retries for each file media set?) before Retrospect will continue with the backup.

 

It doesn't only happen when I start the Engine (I assume the popups are coming from the engine), but it keeps trying to re-connect every 10 minutes or so, bringing the backup to a grinding halt until I notice that the popups are there, and click, click, click, click, click, click, click, click, click, click, click, click. It even does this when there's no backup running at all!

 

Is there any way to tell the engine not to try to connect to the NAS without deleting the file media sets in the Media Sets setup, and then having to rebuild them?

[prl]

The NAS connection seems to be being triggered each time there's a change in the backup state: start of backup, start of verify, start of second optical media set member, start of verify of second member, etc. All, as far as I can see, without any real need to have the NAS connected.

[Don Lee]

Deleting the media sets from the GUI/engine does not require you to rebuild them. When the NAS comes back, you can simply re-add them. I think that's the way to go.

[prl]

I'll give it a go. The behaviour is annoying enough that I'll put up with a rebuild if it's needed.

[prl]

I deleted the media sets as iCompute suggested, and that got rid of the griping about the NAS that I didn't want to use not being there.

 

I got a replacement for the router/NAS today, and I used Locate to put the Media sets back.

 

When I ran one of my backups (No Media Action) to the NAS, I got the "Needs Media" request button. I clicked on the button, and got the file browser to select the media set. Unfortunately, the browser didn't have the NAS in its list of volumes. So I told it to stop. Which, being the well-behaved backup program that it is, it didn't. Just got the SPOD.

 

I tried quitting the Console. No luck. I waited a while and did a Force Quit on the Console, which then exitted. I restarted the Console, but it hung at scanning the third of 7 media sets. But it would at least exit. Tried again, same. Tried stopping the Engine. Instead of stopping, it went to 140% CPU usage, and stuck there. So I did a Force Quit on that. Restarted the Engine, restarted the Console, but the Console hung at the same point. Quit the Console, and tries stopping the Engine. Same 140% CPU. Force quit of the Engine.

 

So I rebooted, got it all together and tried Repairing the Media sets. That ran OK. However, a Repair on the media set shows up as a Rebuild in the activities window!.

 

Tried doing a No Media Action backup again. The backup progressed a bit over half way (in data volume), and then I got the "Needs Media" request button, but although it now shows the NAS folders, when select the media set member folder that it's asking for and I click OK, *nothing happens*.

 

Going to try a media set rebuild.

[Don Lee]

There is no mention above of what kind of media set is on the NAS. The "needs media" on a disk set generally means that either there is no "member", or the "members" are all full. This implies to me that you are using "disk" media sets, and I have found that these *cannot* be moved. When you "bring back" the NAS, you would have to do so without changing the name of the mount point or any other path components. If you did that, I *believe* that you could sti;; just "re-add" the set, but if you don't match up the names, my experience is as you describe.

[prl]

They're Disk media sets. I did "bring back" the NAS with exactly the same path to the members. The single members for the two backup sets were each 30% or less full. They did not run out of space when the backups ran after the rebuild.

 

And even if all of those problems happened, the behaviour of the Engine was not in the least robust. To put it as nicely as possible.

[NTmatter]

As a last-ditch effort, you may want to give the Share (and possibly the NAS itself) a different name. I ran into some similar symptoms after moving a share to a new host.

[prl]

Thanks, but I already have the backups working after doing the Rebuilds.

[zeitkind]

I ran into the same problem. If a NAS-share is not available, it's almost impossible to stop Retrospect (10.5.0 / 145) from triggering such popups. Took me a while to stop the script and quit Retrospect... So, well, IMHO backup to NAS is simply broken, because a non-available NAS-share will hinder all backups. Why not simply put a "try" in an AppleScript?? *sigh*

So I might have to go back to the roots and mount NAS-shares via nfs during system boot? Or is there any workaround?

[MikeHutch]

SOLUTION: Stop Retrospect's long loading time.

 

This is my method:

1) In Media Sets panel: find all your old media sets, that are not currently accessible at the moment and tick the lost button next to each member.

2) In Sources panel: This is the important bit, delete all your old backup destination sources, especially sources that are over a network, such as NAS drives (network hard disks).

 

These actions can usually be easily re-instated, should you need to do a restore, so no harm done.

This will cut down the time taken for Retrospect to load up, by a huge margin.

[zeitkind]

SOLUTION: Stop Retrospect's long loading time.

 

Well, no. The problem matters when a NAS is the target for backups, not a source. Tbh, I'm about to kick Retrospect (use/admin it since v2.0) and use rsync on all sites. If this company is after many years now still unable to fix problems in current versions but instead come out with new versions to make money - no way.