EM: database replication (from VBR) fails [#07640503]

[tinto1970]

good day, I'm struggling in connecting a VBR to the EM, the database replication continue to fail.

Scenario: fresh install on veeam VBR en EM in version 12.3 in two separate VM. Each one has its own MsSql database, 100% dedicated to EM and VBR respectively.

The configuration of a previous VBR 12.2 server has been imported in the new 12.3 VBR server.

*after* succesfully importing the configuration, I tried to add the VBR to the EM, the catalog replication went fine but the DB replication always fail. I tried many times to delete and read, under the suggestions of the support engineer.
It was suspected that some "endpoint protection/ips" was blocking the network connections between the two machines, but it was excluded from the customer.

The last solution proposed by the support engineer is to set the following registry key

SqlStatementTimeout
• Type: REG_DWORD
• Value: 18000
• Decimal
Path MSSQL: HKEY_LOCAL_MACHINE\SOFTWARE\Veeam\Veeam Backup and Replication\DatabaseConfigurations\MsSql

I've set it up but I could not restart the services (important jos are running and I cannot take the risk to break something before 4 days off).

I think this could make sense, because the VMs are pretty slow, due to a importa vCPU overprovisioning on the vSphere infrastructure (I see "ready" times that are often pretty huge).
I hope it works, but I wanted to share this in case someone had same issues and can potentially help with this

Thanks! Happy Easter

[david.domask]

Hi tinto1970,

Thank you for sharing the details on the case and what's been tested so far, and sorry to hear about the troubles.

I can understand the logic on the registry value -- personally, I would expect we get a corresponding SQL timeout exception from Veeam on this as the actual failing message, but perhaps Support saw something in the logs which indicated something similar.

Just out of curiosity, are you adding the VBR hosts by FQDN or the short hostname/IP? I'd be curious if IP worked; I seem to recall when the Enterprise Manager authentication changed (v11 I think?) this could cause issues based on the EM certificate, and testing with IP might be reasonable.

[tinto1970]

good day David, thank you very much for your kind answer. Sorry for the delay in mine but we had some day off for Easter

I'm using the FQDN to add the VBR to EM, but before opening the SR I tried to workaround the issue by using the IP, with the same result.

This morning I could finally reboot the VBR server after changing the registry setting, but this did not solve the issue. I already uploaded the fresh logs to the case.