Here many of our databases are in the TB size range. Vacuums often take several hours to complete, and we have schedule maintenance time for a vacuum full of around an average of 4 hours.
The database I work on started spitting out the following message last night: WARNING: database "ops" must be vacuumed within 9361597 transaction
I ran VACUUM as soon as I got in, but then shortly after, it stopped accepting transactions. I stopped the database and opened single user mode to run VACUUM. So my questions are:
1) How long should I expect this to take (database has about 5 TB of data)?
2) Is there a faster way to make the database be operational again?
Sorry for the multiple emails, but I forgot to mention that I'm using 8.4 on RHEL 6.4.