VSphere replication marked as Not Active

[AleksESD]

Dear All,

I have configured the SRM connection between two sites. I have paired them successfully and configured the VR connection.

Problem is when I set a replication for a certain VM, after wizard is completed, the vsphere replication monitor shows status - Not Active. On the VM tasks and events I found the following warning:

No connection to VR Server: Not responding

I am able to ping/telnet port 80 both of the replication appliances from this VM.

Could you please help me resolve this issue.

Thank you,

Aleksandar Aleksandrov

[admin]

Hi Alexandar,

Replication traffic goes trough ESX server directly to VRS server so are you able to ping VRS server in your target site from ESX server where your VM is running? ESX connects to ports 31031 and 44046.

This KB will give you more details: http://kb.vmware.com/selfservice/microsites/search.do?language=en_US&cmd=displayKC&externalId=100956...

Thanks,

Dzhem

[AleksESD]

Hi Dzhem,

Thank you again for the support.

I am able to ping the VRS server on the target site from the ESXi servers. I have used the link that you have sent to configure the port forwarding on the DR site.

From the primary site, the vCenter, SRM and VRS servers of the DR site, have the same IP,  which is then forwarded to their local IP`s.

I couldn`t do telnet, because I received the following error on ESX server:

-sh: telnet: not found

But when I try from  a windows machine on the same subnet as the ESX server, the telnet on ports 31031 and 44046 works correctly.

Thank you,

Aleksandar Aleksandrov

[admin]

Hi,

What you mean by "vCenter, SRM and VRS servers of the DR site, have the same IP" could you please give us more details on your network setup?

Thanks,

Dzhem

[AleksESD]

Hi,

Bellow is the description of my network setup.

Primary site:

Network: 10.201.27.0 /24

vcenter: 10.201.27.160

VR: 10.201.27.98

DR site:

Network: 10.20.30.0/24

vcenter 10.20.30.21 / 192.168.137.2

VR: 10.20.30.29 / 192.168.137.3

VM with VPN connection to Primary site : 192.168.137.1

The network adapter of the VPN machine is shared among the 192.168.137.0/24 network. vCenter and VR are accessing 10.201.27.0/24 with static route passing trough 192.168.137.1.

After a VPN connection is established, ip 192.168.244.20 is received. On this IP there is a port forwarding for the following ports:

80,443,8095 and 10443 - 192.168.137.2

31031 and 44046 - 192.168.137.3

So, primary site is always connecting to 192.168.244.20 and different ports are redirected to different services.

Please let me know if it is clear the description above.

Thank you,

Aleksandar Aleksandrov

[admin · Accepted solution]

Hi again,

OK here is the problem.

Basically when you pair with DR site both VR servers are exchanging information between each other for example target VRS server. So in this case you end up with following situation:

1) VRS on DR site has different IP address than the Production site

2) Production site ESXi try to connect to VRS IP on DR site without going directly through your port forwarding machine

3) This makes replication to go in Not Active state

Please check the messages in vmkernel.log file located in your source ESXi machine. Here is a useful KB VMware KB: Troubleshooting VMware vSphere Replication issues that could be helpful.

Thanks,

Dzhem

[AleksESD]

Yes. You are right!

Bellow is the log for the replicated VM GID

2015-03-04T11:32:07.292Z cpu37:37002)WARNING: Hbr: 549: Connection failed to 10.20.30.29 (groupID=GID-bbca090f-31de-46bc-8097-4e1f4070bc2a): Timeout

2015-03-04T11:32:07.292Z cpu37:37002)WARNING: Hbr: 4521: Failed to establish connection to [10.20.30.29]:31031(groupID=GID-bbca090f-31de-46bc-8097-4e1f4070bc2a): Timeout

Is there anyway to change 10.20.30.29 with 192.168.244.20 ?

Thank you,

Aleksandar Aleksandrov

:

[admin]

Hi,

Unfortunately you can't solve this problem on a VR level. This have to be solved on a network layer. I'm not a network expert but I think that both sites DR and Production need to have connectivity between each other on a default gateway layer (at least for the replication servers).

Thanks,

Dzhem

[AleksESD]

I have recreated the hole site-to-site VPN and now every machine from site A has connection to every machine in site B and vice verse.  Now the synchronization passes successfully.

Thank you for your help!