DSXchange

The error message suggests that COL_MDL_ORG is of type VarChar.

What command are you using?

Track down whether it's you at fault. For example look for ExecSH as a before/after subroutine, look for DSExecute called with "SH" as the shell argument from your routines. If you eliminate these possibilities you have a stronger case to take back to IBM.

Don't know.

Have you tried it?

My guess would be that it is OK, since the SQL file is simply read to become the SQL Statement property value.

In a User Variables activity change the line terminators (now @FM) to commas, so that you now have a comma-delimited list for the StartLoop activity, and remove the final comma so that you don't have an empty final name. Field(Convert(@FM, ",", ExecCommand.$Output), ",", 1, Count...

I would partition only by col1, and sort by col1, col2, col3. Sorting by col4 doesn't achieve anything, unless to govern the meaning of First or Last in Remove Duplicates stage. For any given value of col1, all values of col1,col2,col3 will occur on the same partition. Only if there are too few dist...

Despite it being a long time, Craig has correctly remembered the "multi-line" or "block" format. The block of statements needs to be terminated with an End statement. You won't be able to run a detached job. I prefer initial capitals to all upper case, though keywords in the BASI...

Depends. Why are you calling it?

An alternative to Craig's most recent suggestion is to make sure that your sub-sequences throw a warning (but do not need to abort). This, too, can be detected upstream (set parent sequences to log a warning if any activity does not finish with a status of OK). If you really want an esoteric solutio...

Perhaps, then, the original SQL's join is supported by indexes, a luxury that DataStage does not have (and the reason that DataStage requires inputs to the Join stage to be sorted on the join key).

A Numeric data type with sufficiently large precision should be adequate. Try dfloat (Double), since this is the largest possible numeric data type, and all numeric data types can be implicitly converted to it. Also try reducing the severity of metadata mismatch reports in any Connector stage that y...

Oracle 12 is not officially supported for the metadata repository from Information Server version 8.1, primarily because Oracle 12 did not exist when 8.1 was released. You are totally reliant on Oracle's promise of backwards compatibility.

Appoorva wrote:Do I have to start any other services(like xmeta) prior to this?

Yes, although it should auto-start on boot.

You must also inspect the log files mentioned in the error message.

If you have any VarChar columns change them to Char.

Could it be that the multiple instances are all (or some of them) trying to access file \\.\pipe\Application-RT_SC274-App_Splunk_Message.CUSTOMER_NATL at the same time? The operating system only allows one writer at a time.