8 Krypton

Index failing on Full VMDK backups

Jump to solution

I use Networker 7.63 mgnt console. I also have a proxy server for image backup jobs. VMware esx 4.1x

The full image backup always completes fine but the index always fails to start. It's the typical warning message about

waiting for a writable volume to backup pool "xxxx" to become available etc.  I've checked the pool/media/parallisem/device settings.

I don't use or want to use different pools for indexes.  Anyone have a suggestion why these indexes don't work?  I'm only running one

job at a time for testing so nothing else is using the device/pool.

Thanks in advance,

SB

Tags (3)
0 Kudos
1 Solution

Accepted Solutions
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

I assume you have pool Images.  That one has volume images.001 on storage node.  Pool Images has your group in pool configuration.  When index is to be saved, which is also part of the group run, it looks for volume from pool Images on backup server, but that one does not exist.  Why backup server?  Because index and bootstrap are owned by backup server and backup server will wrote those on backup server and not storage node.

Two ways to remedy this:

- create volume for Images pool on backup server too or

- create pool which will only save index: and bootstrap save sets and will have volume on server of course

12 Replies
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

I assume you have pool XXXX only on proxy and not on server when index backup happens.  Due to that you get what you get.  Just create one more device and add it to pool XXXX and that's it.

8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

Thank you Hrvoje. I'm looking into that now. I'll let you know what happens.

0 Kudos
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

I tried a few things but no luck. I guess i dont know exactly what you mean Hrvoje. On the media pool what device targets should i have check marked? I have the proxy server device of course. I've also checked Default but that didnt help.  I've made sure the devices are all mounted.

0 Kudos
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

Go to command line of backup server and do nsrmm command and post output here.

8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

adv_file disk Full.001 mounted on \\dd01\backup\exchange, write enabled

adv_file disk images.001.RO mounted on rd=emcproxy1.actransit.lan:\\dd01\backup\images\_AF_readonly, write protected

adv_file disk Full.001.RO mounted on \\dd01\backup\exchange\_AF_readonly, write protected

adv_file disk networker1.actransit.lan.003.RO mounted on \\dd01\backup\windows\_AF_readonly, write protected

adv_file disk images.001 mounted on rd=emcproxy1.actransit.lan:\\dd01\backup\images, write enabled

adv_file disk networker1.actransit.lan.003 mounted on \\dd01\backup\windows, write enabled

adv_file disk networker1.actransit.lan.001 mounted on \\dd01\backup\default, write enabled

adv_file disk networker1.actransit.lan.001.RO mounted on \\dd01\backup\default\_AF_readonly, write protected

Also, I recreated all the proxy items from scratch.  Proxy server, storage node, media pool, device etc.  The first vmdk image backup i did worked as normal but once again the index wouldnt backup. Here's the error message below (FYI... the server CA has *@* access to my Networker server.  Also, the proxy server was setup in datadomain for access.)

ca:index 1 retry attempted

ca:index 39077:save: error, no matching devices for save of client `networker1.actransit.lan'; check storage nodes, devices or pools

ca:index

ca:index 5777:save: Cannot open save session with networker1.actransit.lan

<ERROR> :  Failed with error(s)

0 Kudos
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

I assume you have pool Images.  That one has volume images.001 on storage node.  Pool Images has your group in pool configuration.  When index is to be saved, which is also part of the group run, it looks for volume from pool Images on backup server, but that one does not exist.  Why backup server?  Because index and bootstrap are owned by backup server and backup server will wrote those on backup server and not storage node.

Two ways to remedy this:

- create volume for Images pool on backup server too or

- create pool which will only save index: and bootstrap save sets and will have volume on server of course

8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

Thank you Hrvoje. I understand what you're saying but i'm stuck on creating a new volume for images pool on backup server. How do you do that?

0 Kudos
Highlighted
8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

I have no idea what disk storage do you use, but I assume this is Data Domain due to those dd thingies in name path.

On server you already have:

\\dd01\backup\exchange

\\dd01\backup\windows

\\dd01\backup\default


Now, you can't use \\dd01\backup\images since this is already used by storage node. So, just create folder images 2 so that you have \\dd01\backup\images2 and add it as device (you can copy existing device from backup server and then you just rename path) and as last step just label it to Images pool. If you never did this before, perhaps best way would be to use wizard - in NMC go to Devices section and from there you select add device using wizard option - it will guide you through configuration and it should be easy.

8 Krypton

Re: Index failing on Full VMDK backups

Jump to solution

Thank you for taking the time to explain it so i could easily follow. 

0 Kudos