DSXchange

This isn't about a limit on the number of records you can load, but how many you can bind into a single array. Let's start with how exactly you are 'loading' this data, can you provide the options you are using in the connector? What array size are you using?

As to what the buttons look like, Rewind and Play respectively.

What kind of "better solution" are you looking for?

The only time I've ever seen anything remotely like this was when a slightly different (aka 'buggy') Oracle client version was installed in one environment, hence the question.

Not sure how you get extra anything in a one character field, can you be more specific as to exactly what it is you are seeing?

Is the version of your Oracle client different in production?

g_rkrish wrote:Do any thing need to be done when write and read from same XML.

Wait, wait, wait... are you saying this only happens when you are writing back to the same xml file that you are reading?

Right, in some fashion you'd need to preserve the original file. You could use an After Job routine to check the INTERIM status and, if the job was aborting, swap old back for new. Or, as noted, write to a new name and if the job was going to finish normally, rename the new name to the old name.

In other words, no. The moment the job starts and the stage 'opens' the file is overwritten.

Are all results always zeroes or do some work and some do not? Post your actual syntax, that and confirm for us that the values in your 13,5 will actually fit in a 9,5 decimal.

And seriously, "+ve"? It was too much trouble to spell out the word "positive"?

It's not space or corruption, as a exact search for an appropriate part of that error message would reveal.

You won't be able to use the stage for that. If you really need that functionality, you'll need to roll your own solution - script, routine, etc.

Regardless, you can only get multiple records back from a lookup by using the ODBC or UV stages and by enabling that option on the lookup link itself. Search for "multi-row result set" here for many discussions on the subject.

When you say "same query", how same is same? Are you saying each environment connects to the same Oracle instance and runs the exact same query with the same stage settings and you get wildly different run times? Or each ETL environment sources from a different Oracle instance, dev for dev...

Not sure where that is documented other than in posts here. Unbounded varchar fields are stored as you would expect but declare a size (which most everyone does) and it takes that full size in the dataset.

kogads wrote:I still cant understand how we can track the error file if we have the environment variable $APT_IMPORT_PATTERN_USES_FILESET set to true and File Name column is provides in sequential file?

It lets you know which one of the (possibly many) files you are reading that any particular record is from.