cancel
Showing results for 
Search instead for 
Did you mean: 

VCBMOUNTER - send_keep_alive: sending keep alive mesasage

Santosh_Kalugad
Level 4
I am taking backup using VCB. I am having Netbackup 6.5.2 server which is acting master server, VCB server and Vmware Vi server. When I am trying to take backup I am getting

14:20:51.546 [484.6680] <2> send_keep_alive: sending keep alive message after vcbmounter start to generate snapshot and it is taking 1 hr to generate snapshot of 10 GB VM. Can someone please explain me why it is taking so long time?

5 REPLIES 5

Alasdair_McQuir
Level 4

VMware images can take a long time to create if they have a lot of small files in. It is very dependent on your infrastructure.

 

In general do not run more than one fullvm at a time from the same set of disks as another backup, do not over load your san as this can cause timeouts. 

 

A basic guideline with early esx is to run backups in series rather than parallel. 

 

Try Expanding your backup start notify and client read timeouts, try 7000 or more seconds....

 

Al 

Santosh_Kalugad
Level 4

Hello Alasdair,

 

Thanks a lot for replying to my query.

 

I am taking backup of only one VM at a time. But when I had done POC for VCb that time I used three test VM's with same size and it was taking 5-10 mins to generate snapshot. After that I performed same task in Production and there it is taking around 1 hr to generate snapshot of same size Vm's.

 

Can you please suggest something more on the same.

Anonymous
Not applicable

I have found major differences in timings of VM backups taken based on the options you choose for the command vcbmounter, and this appears to be still restricted in NB 6.5.2

 

If you use NetBackup 6.5.2 they added extra choices within the Snapshot Client Options but have not gone far enough!

 

And this can have a serious impact on the speed of transfer of the snapshot to your staging area on the proxy.

 

Doing backup tests on a VM that is 373GB in size (12GB+361GB VMDK's)

BTW I have a Fibre Channel SAN infrastructure. 2GB HBA's, switches. HP EVA8000 FC SAN disks.

My staging area on the proxy is also a SAN disk, it is RAID0 and is 800GB in size

Using VCB 1.5 Framework

 

On the proxy from the command line I execute

 

vcbMounter.exe -h <vc_server>:443 -u <vcb_user> -p **** -c vmName.cache -a ipaddr:<vm_fqdn> -t fullvm -L 6 -r "E:\mnt\<vm_fqdn>" -m san

 

Ran this twice and both times got it to transfer 373GB in 66 minutes

 

NetBackups turn......

 

Now vcbmounter has defaults that it applies and these are the export flags

 

  <export-flags> := -M (0|1) -F (0|1)

Export Flags:
 -M: If set, the disk is exported into a single (monolithic) file.
     When turned off (default), the disk is split into multiple 2GB files.
 -F: If set, the disk is exported as "flat" disk, with no optimizations.
     When turned off (default), the exported disk files will be more compact as
     unused space in the disk image is not included in the  exported file.

 

In NetBackup there is no option in the Snapshot client options for the -M flag, the snapshot provider somehow sets this flag to 1 as seen when looking at the bpfis log on the proxy. I set the other option to optimize the VM backup.

 

By not allowing access to this flag Netbackup creates a large monolithic file. And guess what the backup time creeps and takes longer.

 

Using NetBackup to control the backup and via its options it takes

 

66 minutes to transfer 230GB to my staging area. !!

 

My optimal configuration and fastest time to backup in my environment seems to be the  defaults of vcbmounter - 2GB chunk files and optimized.

 

With both -M 1 and -F 1 set it took 17 hours to backup 376GB to my staging area!!

 

And of course during this time your VM is writing to the delta file which could have grown to a massive size and given you another headache to worry about on the sizing of your datastores, or even when it tries to commit this delta file into the VM after the backup is complete.

Message Edited by Stuart Green on 08-07-2008 07:01 AM
Message Edited by Stuart Green on 08-07-2008 07:06 AM

Anonymous
Not applicable

At last but have to wait! 6.5.4!

 

NetBackup for VMware slow backup - new published KB article

 

GENERAL ERROR: VMWare Consolidated Backup(VCB) using NetBackup 6.5.x are slower than expected

http://seer.entsupport.symantec.com/docs/311193.htm

 

 

marekkedzierski
Level 6
Partner
hmmm ... so.. bpbkar uses 32k block only with VCB backups (Enterprise Client) or with all file system backups ?