whats wrong with V7.0R2 replication

[pcmmitham]

since I've upgraded to veeam 7.0R2 replication has become an unusable feature! my replication times have increase by 2 - 3 times! yes the new version of veam B&R is SLOWER! backups are about the same (too bad there wasnt an improvement there!) If I could revert to 6.5 I would but I need 7.0R2 in order to connect to a new ESXi 5.5 host and backup a server 2012R2. I have a support ticket open with Veeam (# 00481417) but I'm getting the runabourd with them. their only answer is "something in your infrastructure has changed, its cant be Veeam" the Only thing that changed was my upgrade from 6.5 to 7.0R2 last Wednesday my replications ran as usual. Thursday I upgraded and since then replication is not practical! for example I have a job that replicates 2 DC's to my remote site in version 6.5 this took 25-30 minutes per night, now the same job takes 1 hour, the "read rate" NEVER moves above 1MB/s

We have Riverbeds between the sites and these help tremendously! I even had Riverbed look at the devices while I was replicating and they think that Veeam is now doing some sort of encryption on the data as Riverbed is no longer able to optimize their traffic.

HELP anyone have a solution...Veeam is just ignoring this!

[foggy]

Pierre, could you please describe your setup in a bit more details so we could review it and probably come up with some hints on how to improve performance and possible reasons of the slowdown?

[Gostev]

I can confirm that the replication engine did not have any changes between 6.5 and 7.0R2, and that you are the first to report slow down with over 30% of our customers already on v7. And with 80'000 users, any real product issue immediately results in a few pages long topic on this forums. Based on these facts, I tend to agree with support engineer conclusion. Unless this is some brand new issue introduce in R2 update.

I realize this does not help you, but this is to answer the question you put in the topic name - there is nothing wrong with v7 replication in general.

As for troubleshooting steps, I would try replicating a test VM locally to see if you get faster processing speed.

[pcmmitham]

My setup is as follows

LOCAL
---------------------------------
Vmware environment
vcenter 4.1
3 PE R710 hosts running ESX 4.1
3 Dell equallogic modules
1GB iscsi dedicated network

veeam server server 2008R2 64-bit
PE 2950 2x E5430 2.66 GHz
2 x 1gb iscsi connections to the SAN infrastructure
MD1000 12TB SCSI arttached drive shelf

network ( local)
Cisco 3750 core switch
Riverbed 2050M wan accelerator
Cisco ASA 5510 firewall
20 Mbps Fiber internet connection

REMOTE
-------------------
Vmware
PE T715 2x L5680 proc
200GB ram
ESXi 5.0
12TB DAS

veeam proxy
VM - server 2008
4 procs
4GB RAM
Hot add to the 12TB DAS on the host

network
--------------------
HP procurve 48 port switch
Riverbed CX555 wan Accelerator
ASA 5510 firewall

the veeam jobs are configured to not be optimized per Veeam's recommendations for riverbeds (you guys did a case study on use using these devices)

[foggy]

What about the amount of changes being replicated - has it increased or is it the same, just the time it takes to replicate them has increased? What are the bottleneck stats for this job?