I just wanted to know why some VM backups take longer time as compared to other VM being backup at the same during the assigning of required backup infrastructure resources; See the example below it took more than 1 hour for the required backup infrastructure resources to be assigned and this a VM backup of just 32GB size.
[color=#008040]2/26/2016 4:36:08 AM :: Creating VM snapshot
2/26/2016 4:36:15 AM :: Collecting disk files location data
2/26/2016 4:38:11 AM :: Removing VM snapshot
2/26/2016 4:41:02 AM :: Queued for processing at 2/26/2016 4:41:02 AM
2/26/2016 4:41:02 AM :: Required backup infrastructure resources have been assigned
2/26/2016 5:55:26 AM :: VM processing started at 2/26/2016 5:55:26 AM
2/26/2016 5:55:26 AM :: VM size: 32.0 GB
2/26/2016 5:55:46 AM :: Network traffic will be encrypted
2/26/2016 5:55:48 AM :: Saving [3PAR_SG_08] 994560APDTST02/994560APDTST02.vmx
2/26/2016 5:55:49 AM :: Saving [3PAR_SG_08] 994560APDTST02/994560APDTST02.vmxf
2/26/2016 5:55:50 AM :: Saving [3PAR_SG_08] 994560APDTST02/994560APDTST02.nvram
2/26/2016 5:55:51 AM :: Using backup proxy usaaveeam01.unicef.org for retrieving Hard disk 1 data from storage snapshot
2/26/2016 5:55:57 AM :: Hard disk 1 (32.0 GB) 19.9 GB read at 69 MB/s [CBT]
2/26/2016 6:01:03 AM :: Finalizing
2/26/2016 6:01:07 AM :: Busy: Source 99% > Proxy 13% > Network 2% > Target 0%
2/26/2016 6:01:07 AM :: Primary bottleneck: Source
2/26/2016 6:01:07 AM :: Network traffic verification detected no corrupted blocks
2/26/2016 6:01:07 AM :: Processing finished at 2/26/2016 6:01:07 AM [/color]
Were there any other backup jobs running? If so then how many concurrent tasks are allowed on your proxies and how many concurrent tasks are allowed on you target repository? Please note that one task means one virtual disk.
That might be the reason. If total amount of disks being processed in exceeds 10 or 12 (depending on which proxy/repo is used) your job has to wait for resources become available. Please try to raise those values. Also please note that if you use NBD mode the amount of disks that can be processed simultaneously is limited to 7.
I'm using SAN (direct SAN access) not NBD. After increasing the concurrent tasks allowed on the target repository, the time taken to assign backup infrastructure resources reduced and backups are faster.