August 13, 2002 at 3:59 pm
I am having tons of problems with IMPORTing data from a very large CSV file (64,000,000 records). I keep getting errors like "tempdb full" or "database full" (of the one that I am trying to import to).
I thought that the maximum # of records is around 2 billion or so with SQL Server. I have also tried shrinking the transaction logs in both tempdb and in the database in question and that doesn't help.
Any thoughts?
In addition, I am trying to IMPORT data from two mirrored tables from two other databases into the same database that I am IMPORTing data into. I keep getting a "table lock" error since the databases I am attempting to import from are highly active with many INSERTS (around 100 or so) per minute and they are active 24-7.
Here is a text visual of what I am trying to accomplish:
machine1: database.dbo.table1
machine2a: database.dbo.table1
machine2b: database2.dbo.table1
machine1 => machine2b
[append and then delete records in machine1]
machine2a => machine2b
[append and then delete records in machine2a]
Hope this makes sense.
If any one has some suggestions for accomplishing this goal, I'd really appreciate it.
Thanks in advance!
August 14, 2002 at 10:51 am
Are you doing this in one transaction? You can set the size of the batch to commit.
Steve Jones
http://qa.sqlservercentral.com/columnists/sjones
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
August 15, 2002 at 1:50 pm
This may be overstating the obvious, but, while SQL Server can handle very large databases, you are still limited by the available physical disk space. How big is your hard drive?
Viewing 3 posts - 1 through 2 (of 2 total)
You must be logged in to reply to this topic. Login to reply