DSXchange

It's disk space that's at issue here, not memory space. Each job creates a number of hashed files and directories, as you observed. These should only be deleted by deleting the job. They are the places where run-time metadata are stored. Create a new project on a different file system, and move some...

Yes.

Though I'd be curious about what a negative interval actually means, in a business sense.

There are already some examples in the Parallel forum. Search is your friend. There's a good one by dsguru2b that mimics the functionality of an Ereplace function.

Welcome aboard. Strike 1: posting in the wrong forum Strike 2: doing it just after Andy moved dozens of posts to the correct fora Strike 3: claiming urgency This is an all volunteer site. All posters post as and when they can, with none being paid for doing so. If you need urgent support, sign up wi...

Open Designer. Open a parallel job. Select the Stage Types branch. From the New tool choose new parallel stage then, from the menu presented, choose Build. Note the names of the tabs. Emplace your code there. You can use F1 to get help or read the pertinent chapter of the Parallel Job Advanced Devel...

Try Search - you will find at least two examples.

Bana, feel free to send money for free receipt ! There is a tool called Parameter Manager for DataStage available from the folks who run this site, but I am not sure that it is still available. There is a forum here at DSXchange relating to it. Vincent published a blog on hacking job parameters with...

Welcome aboard. :D

You must preserve the (hash) partitioning of the existing rows in the Data Set. Therefore any new rows appended must be identically hash-partitioned with the existing rows (that is, partitioned on the same key columns).

Check that you are using a compatible C++ compiler.

The parent directory of the orchadmin command must be findable in your command search path (PATH environment variable) or have its complete pathname specified - for example $APT_ORCHHOME/bin/orchadmin Others have posted about the need to have the APT_CONFIG_FILE environment variable set correctly.

The number of player processes has been restricted to two, presumably by the stage in question being executed in a two-node node pool. When run on a configuration containing other than two nodes, this alert message will be generated, solely to make you aware of the mismatch. The job will still run, ...

Depending on whether the source is a timestamp or a string, extract the date portion using a conversion function (with appropriate timestamp format string) then substring. Once you have a string, rearrange it with substring and concatenation, making sure to include leading zeroes where required. The...

BCIGetNext() is the function called to return the next row from a SELECT query made through the ODBC or UV stage type. For some reason this has failed - perhaps the connection to the database server was lost.

Now it's time for nilaytewari to read the properties of the MONTH.TAG Transform, where it explicitly states that a DataStage internal format date is required for the input. I'd bet that this is the magic missing step.

ds_is_fun wrote:What does Job control do?

Anything you like.

In job sequences this tab holds the code generated by compiling the job sequence (in read-only form).

Otherwise you can custom code job control routines to do pretty much anything you can imagine, including make lunch.