Jump to content

media set goes AWOL during verify


kswisher

Recommended Posts

Retrospect Version 8.2.0 (399)

Media sets are on external drives (not a NAS server) connected to the Mac that is running the RS 8.2 server.

 

I am having a problem where retrospect seems to loose/can't find the media set in the middle of backup or media set verify activity. The backup or verify completes partially, then errors out with a

[Media request for "1-RS8 Users Backup Alt" timed out after waiting]

message in the log.

 

In experiments today, I discovered that after an activity fails due to missing media, if I then select any media set and click the "locate and open media set" button, retrospect does not display the disk volume containing the failed media set in the list of volumes. In the mean time, the "missing" volume is still available and accessible via the finder of the host machine. At least once, the volume showed up again in the list after clicking around a bit within retrospect GUI.

 

I seems to happen more often with one of my external drives than others that also contain similar media sets. Is this a new bug on the latest release, or perhaps an early sign of pending disk failure?

 

Link to comment
Share on other sites

It's really important to describe things accurately, to help others know exactly what you are doing and exactly what you are seeing when you do it.

 

Media sets are on external drives...connected to the Mac that is running the RS 8.2 server.

 

A Media Set is the "set" of Catalog File and Media Members. With the exception of a File Media Set, all the necessary components of a Media Set would not normally be housed together on external drive(s).

 

http://en.wikipedia.org/wiki/Set_(mathematics)

"A set is a collection of distinct objects, considered as an object in its own right."

 

If you are using Disk Media Sets with Members housed on directly attached external volumes, the Catalog file (to complete the "set) would normally be stored on the engine host machine's boot drive (doesn't have to be, that's just the normally used default).

 

 

retrospect seems to loose/can't find the media set in the middle of backup or media set verify activity

 

So this is where knowing exactly what's happening is necessary. Is Retrospect loosing track of the Member of the Media Set?

 

[Media request for "1-RS8 Users Backup Alt" timed out after waiting]

 

Media requests are generally associated with a change in Members; either adding a new one during a Backup or providing an additional one during a Verify.

 

Does this Disk Media Set have more then a single Member?

 

 

I discovered that after an activity fails due to missing media, if I then select any media set and click the "locate and open media set" button, retrospect does not display the disk volume containing the failed media set in the list of volumes.

 

Do the volumes in question show in the Sources window?

 

Directly attached volumes will show in Retrospect dialogs without user intervention (a different issue then the mounting problem with network volumes). If a directly attached volume (Connected how? Connected to what model of Macintosh?) doesn't display in the Sources window but does show in the Finder, that would be odd.

 

What does the "Mount" command from Terminal show when the machine is in this state?

Link to comment
Share on other sites

It's really important to describe things accurately, to help others know exactly what you are doing and exactly what you are seeing when you do it.

 

Media sets are on external drives...connected to the Mac that is running the RS 8.2 server.

 

A Media Set is the "set" of Catalog File and Media Members. With the exception of a File Media Set, all the necessary components of a Media Set would not normally be housed together on external drive(s).

 

http://en.wikipedia.org/wiki/Set_(mathematics)

"A set is a collection of distinct objects, considered as an object in its own right."

 

If you are using Disk Media Sets with Members housed on directly attached external volumes, the Catalog file (to complete the "set) would normally be stored on the engine host machine's boot drive (doesn't have to be, that's just the normally used default).

 

Ok, this is the case. Catalogs are in the default location in /Library/Application Support/Retrospect/

and the media set members are on external Disks with location /Volumes/backup/ or /Volumes/backup2/

 

Each media set has only one member, although there multiple media set members per disk. 200+ GB of free disk space and no single member is over 180 GB and most are half that size.

 

retrospect seems to loose/can't find the media set in the middle of backup or media set verify activity

 

So this is where knowing exactly what's happening is necessary. Is Retrospect loosing track of the Member of the Media Set?

 

Yes. RS looses track of the member.

 

[Media request for "1-RS8 Users Backup Alt" timed out after waiting]

 

Media requests are generally associated with a change in Members; either adding a new one during a Backup or providing an additional one during a Verify.

 

Agreed. This is the normal behavior. In my case, Retrospect finds the media member, starts reading or writing to the member, then suddenly can't find the media member a few minutes into the operation.

 

Does this Disk Media Set have more then a single Member?

 

No, only one member per media set.

 

I discovered that after an activity fails due to missing media, if I then select any media set and click the "locate and open media set" button, retrospect does not display the disk volume containing the failed media set in the list of volumes.

 

Do the volumes in question show in the Sources window?

 

No, these volumes are never in the sources window. All I see here is the DVD drive.

 

Immediately after a failure, If I select a media set member in the GUI, and click Locate button, the volume "Backup" does not show in the list when it asks you to "Select the Catalog" but the other volumes still show.

 

Directly attached volumes will show in Retrospect dialogs without user intervention (a different issue then the mounting problem with network volumes). If a directly attached volume (Connected how? Connected to what model of Macintosh?) doesn't display in the Sources window but does show in the Finder, that would be odd.

 

Yes, I thought this was a different issue than the mounting problem with network volumes.

 

Late 2007 iMac running 10.6.5 desktop OS, 4 GB RAM, and over 600 GB of free disk space. Three powered 1 TB external drives, only used for backup, connected via daisy chained FW 800. I have swapped the power bricks and cables around with no change in behavior. Suspect a flakey drive or interface board... but how to test?

 

What does the "Mount" command from Terminal show when the machine is in this state?

 

After failure...

 

iDad:~ kswisher$ mount

/dev/disk0s2 on / (hfs, local, journaled)

devfs on /dev (devfs, local, nobrowse)

map -hosts on /net (autofs, nosuid, automounted, nobrowse)

map auto_home on /home (autofs, automounted, nobrowse)

/dev/disk2s3 on /Volumes/Backup (hfs, local, nodev, nosuid, journaled, noowners)

/dev/disk3s2 on /Volumes/Backup2 (hfs, local, nodev, nosuid, journaled, noowners)

/dev/disk1s10 on /Volumes/Clone_Disk (hfs, local, nodev, nosuid, journaled)

vfstool@fuse0 on /private/tmp/12359/C (fusefs, nodev, noexec, nosuid, synchronous, mounted by kswisher)

afp_0V1wtE0Ytrkq00mlHE0Pyetl-1.2d000040 on /Volumes/TM Data (afpfs, nobrowse)

/dev/disk5s2 on /Volumes/Backup of iDad (hfs, local, nodev, nosuid, journaled)

iDad:~ kswisher$

 

TIme machine happened to be running when I ran this.

 

After a few minutes time, the disk is showing again when I select a media set member in the GUI, and click Locate button.

Link to comment
Share on other sites

Have you tried a firewire hub instead of daisy chaining the drives? I have heard that daisy chaining can sometimes cause problems especially if the the drives are not set up for daisy chaining. The Finder may see the daisy chain fine but Retrospect may not so it would be interesting to know if the problem goes away when the drives are not daisy chained.

Link to comment
Share on other sites

I tried an experiment and disconnected all but the one external drive that tends to exhibit the problem most often. Did a simultaneous verify of the 5 media sets that had members that reside on this one external drive. It took longer than normal, but still failed. All 5 media set verifications failed with the

"Media request for timed out after waiting" error.

Link to comment
Share on other sites

I asked:

[color:purple]"Do the volumes in question show in the Sources window?"[/color]

 

You replied:

[color:green]"No, these volumes are never in the sources window. All I see here is the DVD drive.[/color] "

 

All you see where is the DVD drive?

 

You don't see the internal hard drive in the Sources window?

 

That would be wrong, of course. You should see all the /dev/disks connected to the Engine Host machine in Sources.

Link to comment
Share on other sites

I asked:

[color:purple]"Do the volumes in question show in the Sources window?"[/color]

 

You replied:

[color:green]"No, these volumes are never in the sources window. All I see here is the DVD drive.[/color] "

 

All you see where is the DVD drive?

 

You don't see the internal hard drive in the Sources window?

 

That would be wrong, of course. You should see all the /dev/disks connected to the Engine Host machine in Sources.

 

Dhoh! My mistake. I only see the DVD drive in the Storage Devices. I was thinking of the external drives as storage devices (what I use them for), not sources.

 

In sources, I see all the internal and external hard drives, as well as the remote clients and their drives. When the failure occurs, the sources window shows the AWOL drive with a ~ in front of it's name instead of a radio button.

Link to comment
Share on other sites

When the failure occurs, the sources window shows the AWOL drive with a ~ in front of it's name instead of a radio button.

 

So this tells us Retrospect is not loosing track of the Member per se, it's loosing track of the Volume on which that Member is housed.

 

Late 2007 iMac running 10.6.5 desktop OS, 4 GB RAM, and over 600 GB of free disk space. Three powered 1 TB external drives, only used for backup, connected via daisy chained FW 800. I have swapped the power bricks and cables around with no change in behavior. Suspect a flakey drive or interface board... but how to test?

 

Certainly sounds like a hardware problem. You've got three FW cables and four FW bus connections, any of which could be a problem.

 

The obvious test would be to connect one drive at a time (starting with the one that fails most often); each/any time there's a failure, change _one_ component and test again. If you're lucky it might turn out to be one of the cables; if you're unlucky it might turn out to be FW on the iMac.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
×
×
  • Create New...