M
MarkH
We're using ADO.NET and SQl Server 2K.
I'm updating an integer column in a seperate table which I use to hold
sequence numbers for different reference strings.
Together the string and int form a unique code which is we use as user
reference for invoices, orders etc, (the PKs are hidden).
Before I enter a new invoice we:
1) open a transaction in ADO.NET,
2) call a stored proc to increment the relevant sequence and return the
new value,
3) insert data into the invoice tables (including the unique code from
step 2)
4) commit / rollback the transaction
At the moment we're using the default Isolation
Level, is this safe, or do we need to increase it?
I'm updating an integer column in a seperate table which I use to hold
sequence numbers for different reference strings.
Together the string and int form a unique code which is we use as user
reference for invoices, orders etc, (the PKs are hidden).
Before I enter a new invoice we:
1) open a transaction in ADO.NET,
2) call a stored proc to increment the relevant sequence and return the
new value,
3) insert data into the invoice tables (including the unique code from
step 2)
4) commit / rollback the transaction
At the moment we're using the default Isolation
Level, is this safe, or do we need to increase it?