Odd error during database migration to Postgres

[William Bohn]

I created a ticket on this today but also figured it may be something that is known by the community more than the Veeam team so covering both basis.

Issue: During the migrate wizard at the 'Target Database' stage, we get the error 'no pg_hba.conf entry for host "mac address XXX" user "<our domain>\backupuser", database "<our domain>\backupuser", no encryption

I get that same error whether I pick Windows authentication or try to use the postgres superuser with Native authentication

I installed the Postgres DB that came with the v12 ISO onto the same server that is running Veeam B&R.
I didn't create any extra users or databases with postgres. I used the same admin level backupuser account to install Postgres that we use to run Veeam B&R.

Why would it even think there is a database named the same as the domain\user ?
Why would it throw the same error looking for the domain\user if I choose to use Native authentication?

If anyone had a trouble free Windows install and migration to postgres, can you say if you created different accounts and databases to iron this out or did you do as little as I did but didn't get these errors?

thanks for any two-cents that others can toss into this situation!

[Mildur]

Hi William

mac address XXX

What you see is most likely the IPv6 address of your server. Not the mac address.
If you use the hostname of the server for the PostgreSQL connection, windows will connect with its own ip address. And PostgreSQL doesn‘t allow this connection by default without editing the pg_hba.conf file.

Please try 127.0.0.1 or localhost as the server address for the PostgreSQL connection. That will force the operating system to use an internal connection to connect to the PostgreSQL server. Such connections are allowed by default.

Let me know if it works that way.

Best,
Fabian

[William Bohn]

Thanks for the tip - it was odd since it was populated by default to use the server name as host.. changing it to localhost and also using the postgres super user allowed the migration to complete!

I have one other question, I ran the suggested script:

Set-VBRPSQLDatabaseServerLimits -DumpToFile c:\pgres\dump\pfile
WARNING: SQL Configuration script was dumped to file 'c:\pgres\dump\pfile'. Run it and restart postgres service manually to apply changes. Make sure there are no
running jobs before restart.

but confused if it already made the changes and waiting for restart or as it says, that I have to run all those ALTER SYSTEM SET configs?
Do I add these into a postgres configuration file or run it somewhere in postgres using pgAdmin? (which database/schema to run it as, etc?)

It would be nice if the link that suggested this went the last mile to just show the end result steps. The link suggests just running Set-VBRPSQLDatabaseServerLimits and then nothing else.

https://helpcenter.veeam.com/docs/backu ... ml?ver=120

[CALisovich]

Thank you for posting this! I had this error while upgrading VBR to 12.1 and changing the server address to localhost for the PostgresSQL server resolved my issue.

[tthomas1@ebsco.com]

In case anyone searches for this who is also trying to migrate the DB on Enterprise Manager, the equivalent step is to add /servername:localhost to the Veeam.EM.DB.Migration restore command.