Discussions specific to the VMware vSphere hypervisor
Post Reply
ddejarnett
Influencer
Posts: 14
Liked: 1 time
Joined: Aug 08, 2016 4:13 pm
Contact:

First WAN Replication Questions

Post by ddejarnett » Aug 08, 2016 5:26 pm

I've been using Veeam for a while with onsite backups and it runs awesome. My office has now setup a remote site and we want to create replicas there for DR. To reduce the replication over the WAN, I ran a backup of my local VMs and transported it to the DR site. When I attempted to seed my backups on the first run, I didn't realize that my excluded disks would cause a problem, so instead I restored the images and then mapped the replicas from my local site to the DR site.

I ran my first "mapped" job yesterday and I think it's running slow, but this is my first WAN job, so I'm not sure. Below is the log from one of my VMs. Does this look normal?

Code: Select all

8/7/2016 9:51:48 PM :: Queued for processing at 8/7/2016 9:51:48 PM 
8/7/2016 9:51:50 PM :: Required backup infrastructure resources have been assigned 
8/7/2016 9:51:53 PM :: VM processing started at 8/7/2016 9:51:53 PM 
8/7/2016 9:51:53 PM :: VM size: 160.0 GB 
8/7/2016 9:51:56 PM :: Discovering replica VM 
8/7/2016 9:52:05 PM :: Getting VM info from vSphere 
8/7/2016 9:52:11 PM :: Creating VM snapshot 
8/7/2016 9:52:14 PM :: Using source WAN Accelerator D9-VEEAM 
8/7/2016 9:52:14 PM :: Using target WAN Accelerator 192.168.2.2 
8/7/2016 9:54:19 PM :: Preparing replica VM 
8/7/2016 9:54:20 PM :: Processing configuration 
8/7/2016 9:54:34 PM :: Cannot find dirty block information in the previous restore point, CBT will not be used 
8/7/2016 9:54:34 PM :: Creating helper snapshot 
8/7/2016 9:54:45 PM :: Using source proxy VMware Backup Proxy for disk Hard disk 1 [nbd] 
8/7/2016 9:54:45 PM :: Using target proxy 192.168.2.2 for disk Hard disk 1 [hotadd] 
8/7/2016 9:55:19 PM :: Hard disk 1 (160.0 GB) 58.6 GB read at 22 MB/s
8/7/2016 9:56:06 PM :: Global cache synchronization completed. 
8/8/2016 2:51:31 AM :: 10.8 GB transferred over network, 25.5 GB obtained from global cache, 9.5 GB obtained from target repository 
8/8/2016 2:52:11 AM :: Removing VM snapshot 
8/8/2016 2:52:55 AM :: Deleting helper snapshot 
8/8/2016 5:11:29 AM :: Finalizing 
8/8/2016 5:11:49 AM :: Busy: Source 14% > Source WAN 85% > Network 43% > Target WAN 100% > Target 81% 
8/8/2016 5:11:49 AM :: Primary bottleneck: Target WAN 
8/8/2016 5:11:49 AM :: Network traffic verification detected no corrupted blocks 
8/8/2016 5:11:49 AM :: Processing finished at 8/8/2016 5:11:49 AM 
Thanks for any help you can offer a newbie!

David

foggy
Veeam Software
Posts: 17931
Liked: 1512 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: First WAN Replication Questions

Post by foggy » Aug 09, 2016 2:27 pm

Hi David, looks fine to me. WAN acceleration is optimized to reduce the amount of transferred data and is effectively trading disk I/O for bandwidth savings, so disk I/O performance on the target repository is a typical bottleneck for WAN accelerated jobs and faster disks are typically recommended for the target WAN cache.

ddejarnett
Influencer
Posts: 14
Liked: 1 time
Joined: Aug 08, 2016 4:13 pm
Contact:

Re: First WAN Replication Questions

Post by ddejarnett » Aug 09, 2016 4:41 pm

Alex,

Thanks for your reply. Once this first job completes, will it always take this long? The machine in the example log above is the smallest of 7 VMs that I'm attempting to replicate.

Also, in your opinion, is it better to run backups to offsite locations or replicas?

Thanks again!
David

foggy
Veeam Software
Posts: 17931
Liked: 1512 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: First WAN Replication Questions

Post by foggy » Aug 10, 2016 11:51 am

Subsequent runs are typically faster than the initial one, performed right after mapping, since only daily changes will need to be synced. Also, they should use CBT (this one didn't, according to the log).
ddejarnett wrote:Also, in your opinion, is it better to run backups to offsite locations or replicas?
This depends on your needs, replicas allow for faster failback in case of DR, but have less retention and require more space, since are stored in native format.

ddejarnett
Influencer
Posts: 14
Liked: 1 time
Joined: Aug 08, 2016 4:13 pm
Contact:

Re: First WAN Replication Questions

Post by ddejarnett » Aug 10, 2016 1:05 pm

Thanks Alex. I have one last question.

Which type of backup strategy will transfer the least amount of data over the WAN yet still giving DR functionality? A local backup job and then a backup copy job to the remote site OR a replication job to the remote site?

foggy
Veeam Software
Posts: 17931
Liked: 1512 times
Joined: Jul 11, 2011 10:22 am
Full Name: Alexander Fogelson
Contact:

Re: First WAN Replication Questions

Post by foggy » Aug 10, 2016 1:07 pm

Since the number of changes inside a VM in both cases is the same (provided both jobs run daily, for example), the amount of transferred data will also be similar.

pkelly_sts
Expert
Posts: 568
Liked: 62 times
Joined: Jun 13, 2013 10:08 am
Full Name: Paul Kelly
Contact:

Re: First WAN Replication Questions

Post by pkelly_sts » Aug 11, 2016 3:54 pm

Bear in mind that you can do a backup copy to the remote site, then generate remote replicas from that backup copy, so you have both a backup copy AND a replica of your vms, for only one copy operation over the WAN.

Post Reply

Who is online

Users browsing this forum: No registered users and 11 guests