I've just had a case where something happened to an SQL server/database and the transaction log backup detected the databases as "deleted".
Found previously backed up SQL Server transaction log for Instance\DBName
Skipping deleted databases: Instance\DBName
This didn't result in a warning and so the next backups ran with success although it was missing the database.
Wouldn't a one time warning be better in that case, so that customers notice that something has changed?
Thanks for feedback, we will see if that behavior can\should be altered as we are trying to minimize disruptive messages in our logs - what if DB was removed intentionally, say a testing\development one? In that case customer is aware of it, but will still get annoying notifications from Veeam backup jobs?
However, if database is gone in production and customer is unaware, they have a bigger problem than a color of a backup status message.
I do understand that; normally the message/warning should be unnecessary.
(I also know some cases in the past where changes in the notification/warning behaviour inside Veeam caused annoying messages )
In that special case the customer made changes to the SQL server and prepared an update; they didn't delete the database, but perhaps it was set offline.
Then they waited for the next transaction log backup, checked if it ran successfully and afterwards started with the update.
Later they've noticed that the log backup didn't include data/logs.
I agree the need is real, however I would argue this is purely a monitoring product function to monitor the availability of SQL Server databases and alert if something goes wrong with them.
Using backup product for SQL Server database availability monitoring to me is best compared to using your end users for Active Directory monitoring: if their logons start failing and they start placing help desk calls, then this will signal you that your domain controllers are offline. This is really no different from Veeam issuing warnings when some database is found to be offline? And in both cases, you should probably have reacted to the issue waaay before you received those first calls from users (or a warning from Veeam).
I'm also immediately thinking of those dev/test environments many of our customers back up with Veeam. There, databases come and go all the time, which will result in literally every Veeam backup job finishing with the warning.
I'm with you Anton. In almost every case the message/warning would be unnecessary or redundant; this case was rather special because of the planed maintenance and should rarely repeat.
I'll discuss that with our customer,thanks!
)