Importing 19 million records - Primary Filegroup full

[Mongr1l]

I'm trying to import into a db a file that is 19 million records big.

Each record consists of 20 fields, each nvarchar of 20 in length.

I have 115 gigs free space on the drive for the data file, and 115 gigs on the log drive.

The log and data are on separate drives.

When I DTS the file into the DB, at the end of the transfer, after it has moved all 19 million records, it then gives me an error message of "Primary File Group full".

Notice that it doesn't do this in the middle of the transfer, but at the very end of it.

I am told that the problem is the tempDB, that it isn't expanding fast enough.

How do I get this to work?


mongr1l

 

[jdemmi]

Can you constrain the query and load it in smaller chunks?

-- Jason
"It's Just Ones and Zeros

 

[Catadmin]

Well, first see if TempDB is on it's own logical drive, separate from all other DBs. If not, use ALTER DATABASE to move it to one if you have one available with a decent amount of space.

Then check the hard drive space of the drive you have your user DB on to make sure you have a decent amount there.

Then right click the TempDB (and then your UserDB) and change your growth properties for the data file(s). Make sure it isn't constrained to a specific size (some DBAs check the option to make sure a DB can never grow larger than X amount) and increase the percentage / # of MBs that your DBs grow by.

Once you've done importing your rows, if this was a one time only thing, change all your growth values back to their original settings. This way you don't get overgrowth happening when you don't need it.

Hope this helps.



Catadmin - MCDBA, MCSA
"No, no. Yes. No, I tried that. Yes, both ways. No, I don't know. No again. Are there any more questions?"
-- Xena, "Been There, Done That"