Yes and the limit of a direct piece of SQL in BTEQ/Queryman is 1Mb and the limit on a full sProc 64Mb. However when I paste my SQL Statement into a text document it reads 40Kb on the surface. Does Teradata think it's bigger under the covers? How can we see this?
We pasted it to a sProc of its own and it compiles fine. This doesn't tally up with the 64Kb limit. This doesn't help our developers understand the issue any better.
We are running in Teradata Mode. We are using SQL Assistant 7.0.0. There are something like 620 columns in the insert statement.
And why are there limits on sProc size and SQL Statement size anyway?
Thanks, Excuse my ignorance but what exactly are the USING data & and parcel overhead? Can we determine the size of this header info? So really what we are actually left with is less than 64Kb for a SQL Statement within stored procedures. SQLServer & Oracle don't have this limit. If you can have over 2000 columns on a table, there should be no limit in the complexity you can apply to those columns when Querying the table.
Using data is all of the variable data that it has to submit along with your request text. If it is for an insert it would include the data for that insert.
Everything sent to the database is in the form of parcels. A parcel consists of a header and a body. The header says what the body is and its length. The body is the data, which in your case would be the request text.
See Database Administration documentation for more information. Do a search on parcel.