Replication Stats

[homerjnick]

Can someone have a look at my replication stats.

Replicating to my DR site...

Production Site has:

IBM N Series 3600 FC SAN
3 x ESXi5 hosts IBM Bladecenter
Veeam B&R 6 latest patch Windows 2008R2 server on an IBM Blade

DR Site:

IBM DS4300 SAN
2 x ESXi5 on IBM Blades
VM as Veeam Proxy

Site is connected via 10 meg line.

Backed up a 20GB VM at my Production site, transported it to my DR site via USB drive, added it to my repository at the DR site and set replication using it as the seed.

I understand that the first stage is create a VM replica from the seeded backup so stage 1 in the replication has this:

23/05/2012 09:35:08 :: Queued for processing at 23/05/2012 09:35:08
23/05/2012 09:35:08 :: Required resources have been assigned
23/05/2012 09:35:10 :: VM processing started at 23/05/2012 09:35:10
23/05/2012 09:35:10 :: VM size: 14.0 GB
23/05/2012 09:35:18 :: Using target proxy <DR VEEAM PROXY> [hotadd;nbd]
23/05/2012 09:35:19 :: Discovering replica VM
23/05/2012 09:35:43 :: Processing configuration
23/05/2012 09:36:53 :: Creating helper snapshot
23/05/2012 09:36:58 :: Hard Disk 1 (10.0 GB)
23/05/2012 10:20:19 :: Hard Disk 2 (4.0 GB)
23/05/2012 10:38:44 :: Deleting helper snapshot
23/05/2012 10:39:04 :: Finalizing
23/05/2012 10:39:04 :: Applying retention policy
23/05/2012 10:39:16 :: Busy: Source 27% > Proxy 7% > Network 0% > Target 99%
23/05/2012 10:39:16 :: Primary bottleneck: Target
23/05/2012 10:39:16 :: Processing finished at 23/05/2012 10:39:16

Target is the bottleneck at 99% as I would expect since it is transporting to the seeded backup to my DR SAN?

Then the second stage is to copy the changes from the production site to the DR site I assume.

The stats are

23/05/2012 10:39:19 :: Queued for processing at 23/05/2012 10:39:19
23/05/2012 10:39:20 :: Required resources have been assigned
23/05/2012 10:39:23 :: VM processing started at 23/05/2012 10:39:21
23/05/2012 10:39:23 :: VM size: 14.0 GB
23/05/2012 10:39:26 :: Using source proxy VMware Backup Proxy [san;nbd]
23/05/2012 10:39:27 :: Using target proxy <DR VEEAM PROXY> [hotadd;nbd]
23/05/2012 10:39:29 :: Discovering replica VM
23/05/2012 10:39:29 :: Preparing replica VM
23/05/2012 10:39:55 :: Processing configuration
23/05/2012 10:40:20 :: Creating helper snapshot
23/05/2012 10:40:27 :: Calculating digests Hard disk 1 (10.0 GB)
23/05/2012 10:43:29 :: Hard Disk 1 (10.0 GB) 683.5MB read at 123KB/s [CBT]

This is still on going after 2 hours.

Bottleneck stats are:

Busy: Source 18% > Proxy 5% > Network 0% > Target 99%

The two questions I ask are:

Why is the target at 99%, I would have assumed network would be at 99% if the changes are being transported from production to DR?
Why is the HARD DISK 1 read rate using CBT so low?

[Gostev]

1. Not sure. Bottleneck stats you have posted make no sense to me.
2. We cannot read the data from source faster than target can accept the processed data.

[homerjnick]

Guess I'll open a case then...thanks.