Reduce/Eliminate Downtime due to Snapshots

[gggggggg]

We have a pretty decent environment for our size. But not big enough to have any kind of hot DR storage

• VMware standard (6.5) with 4 hosts, and 100 VMs

• Veeam B&R 9.5 on a physical host - 64GB RAM, local storage for Veeam repositories

• All 10GB network throughout

• HP 3Par SAN

A typical backup looks like this:

Code: Select all

9/08/2018 7:01:21 PM :: Queued for processing at 9/08/2018 7:01:21 PM  
9/08/2018 7:01:22 PM :: Required backup infrastructure resources have been assigned  
9/08/2018 7:01:27 PM :: VM processing started at 9/08/2018 7:01:27 PM  
9/08/2018 7:01:27 PM :: VM size: 10.0 TB (8.4 TB used)  
9/08/2018 7:01:27 PM :: Getting VM info from vSphere  
9/08/2018 7:01:34 PM :: Creating VM snapshot  
9/08/2018 7:01:46 PM :: Saving [vmfs-prod-01] vm-filesvr-01/vm-filesvr-01.vmx  
9/08/2018 7:01:46 PM :: Saving [vmfs-prod-01] vm-filesvr-01/vm-filesvr-01.nvram  
9/08/2018 7:01:46 PM :: Using backup proxy vm-veeamprxy-03 for disk Hard disk 3 [hotadd]  
9/08/2018 7:01:46 PM :: Using backup proxy vm-veeamprxy-02 for disk Hard disk 1 [hotadd]  
9/08/2018 7:01:46 PM :: Using backup proxy vm-veeamvvol-02 for disk Hard disk 2 [hotadd]  
9/08/2018 7:01:46 PM :: Using backup proxy vm-veeamprxy-01 for disk Hard disk 4 [hotadd]  
9/08/2018 7:02:29 PM :: Hard disk 1 (100.0 GB) 7.8 GB read at 378 MB/s [CBT] 
9/08/2018 7:02:29 PM :: Hard disk 4 (20.0 GB) 0.0 B read at 0 KB/s [CBT] 
9/08/2018 7:02:48 PM :: Hard disk 2 (4.9 TB) 14.6 GB read at 334 MB/s [CBT] 
9/08/2018 7:02:48 PM :: Hard disk 3 (5.0 TB) 713.0 MB read at 365 MB/s [CBT] 
9/08/2018 7:06:09 PM :: Removing VM snapshot  
9/08/2018 7:06:36 PM :: Finalizing  
9/08/2018 7:06:41 PM :: Deleted file blocks skipped: 2.0 GB  
9/08/2018 7:06:44 PM :: Busy: Source 88% > Proxy 46% > Network 36% > Target 3%  
9/08/2018 7:06:44 PM :: Primary bottleneck: Source  
9/08/2018 7:06:44 PM :: Network traffic verification detected no corrupted blocks  
9/08/2018 7:06:44 PM :: Processing finished at 9/08/2018 7:06:44 PM  

My issue is the outage for the snapshot time. We would love to do 2 hourly backups, but there is a small outage (reading the logs above, I think this example a 30 second outage of the server) during the snapshot stage at the end of the backup. This means I cant do any backups during the day, but my night time backups that disrupt server<->server connections can also cause issues like app to database. This example server is 10TB, but it doesn't matter if its also a 100GB server.

[*]Is this normal?
[*]Is there a way to eliminate it?
[*]How does everyone else manage this?


thanks

[gggggggg]

Sorry, the actual outage might be when the snapshot is taken at the start of the backup, not the end.

[foggy]

Hi Greg, short stun at the time of VM snapshot removal is expected, if this is what you're seeing. To eliminate that, consider using backup from storage snapshots, particularly as your storage does integrate with Veeam B&R natively.