Netbackup for vms from VCENTER vs from VMs , Questions

[SamiEMC]

Hello

We are testing netbackup to get daily ( incremental ) Backup for the VMs configuring the Vcenter

but feedback is :

1) the Admin says it seems slower than normal backup compared for the servers before conversion , is that correct ? is that because vmdk is used here ?

2) the VMWARE admin says it takes the same size everytime when making snapshot , that shouldnt be correct , right ? and if so , they will be using same datastore->may cause problems if full , that will make it full

3) if we install agent on the vms , that will be taking drives backup , will that cause any problems if we do restore for example ?

4) I have some vm which finishes the Backup but snapshot isnt deleted , i will check that again today to see, is this a common issue for some hosts ? the vm is with raw map devices , shall we exclude or change something when having backup for that vm ? or thats not rela

5) What is the main differences between this way and having agents on the vms ? management wise ? or vmdk restore issue ? or what exactly

Please advise

Thanks

[V4]

Hi SamiEMC

Was backup speedy before conversion (pls elaborate). If backup speed is not satisfactory, switch transport method to SAN

Could you elaborate more about  policy and NBU+VMware environment for us to help you in better way

Symantec recommends only to install agent on guest VMs if we need to leverage Application level GRT and have to restore files/folder on Original VMs. For Backup NetBackup+VMware uses vStorage API to read and see content from vmdk, so if you backup vmdk, still you can leverage V-ray feature which lets you restore individual file object from vmdk backup without having to stage your complete backup as standby VM

If snapshots aren't deleted after successful backup, Try performing another backup after some time interval. As it requires stun cycle to remove those snapshots. However its VMWare who handles this snapshot removal process not NBU.

NBU would just request for snapshot (to back it up) and post completion would inform to removal. Later it's  vCenter how it deals with removal depending upon any other criteria which happens in backend etc. There must be event logged for those orphaned snapshots. Check if it helps or post bpfis logs from VMware access host to review of these stall snapshots

As stated in previous para, it's upto your requirement the way you want it . Symantec offers various ways to achieve it. and I'd say it's flexibility which you get . Single pass backup, Sees and Catalogs all data within vmdk, restore individual files, drives,  or complete VMs... Isn't that gr8 

hope it helps. Let us know if you have any further queries

[muhanad_daher]

 the VMWARE admin says it takes the same size everytime when making snapshot.

i have the same problem, and i don't know is that related to configuration in netbackup policy or VCenter.

can you please advise?

[Anonymous]

Make sure you have Change Block Tracking and Block Incremental Backups set enabled. To get the best and optimized from your backup of the VM's at block level. Otherwise you could be still backing up all blocks.