DSXchange

Hi,
Number of ade files are created by RTI Service Job's.because of these file's,tmp direcoty is getting filled.
My questions is,wat happens if i delete all ade files in that directory.i mean if i run the RTI Job again,whether those job's will run or not.Whether is their any other impact on jobs.Wat's the excat purpose of these ade files.
Apprecite your answers.
Thanks,
har

You can certainly delete any old ones but I wouldn't suggest doing so for any running jobs.

I can delete the file's,once the job finsh's.
But wats excat pupose of these files or wat kind of informations does these files contains.

You'd have to call IBM support and ask them that, heck if I know. Or perhaps wait for Mr Ostic to stumble across this thread. You'd have a better chance of that if you posted this question in the proper forum - there *is* one specific to RTI / SOA questions, ya know.

craig..!

The "ade" part is like the DataStage "signature" - if you execute an ipcs command at UNIX you will encounted a number of shared memory segments whose signatures begin 0xade...

Thus the "ade" files for the RTI interface are temporary files needed for storage or communication between processes. Try ls -l to see whether any of them is a special type of file (I'm thinking fifo, or named pipe, here).

In any case, once the job has finished they should be safe to delete. Beware, though, that some RTI jobs are implemented as "always running" jobs.

They're just 'regular' files associated with the RTI Input stages in the jobs. A 'file' command on H-PUX tells me that they are s500 assembler program text but cat'ing one out shows me the xml packets used when calling the service.

Also interesting to note is that we have six 'always on' jobs running right now but only 3 ade files in /tmp right now. Well, 4 if you count the one from a 'regular' job because it has inter-process row buffering turned on between two Transformer stages.

When the jobs are shut down 'gracefully', i.e. from the RTI Console for example, those ade.* files should be deleted. Is that not happening in your case?

Yes,ade files are not deleteting and running out of memmory.
So jobs are getting abort very oftenly.Currenlty we're deleting these file manually.
Is their any automated way or their any better way to delete these file automatically,once job ran sucessfully.

A question for Support, I would think, as that's definitely not polite behaviour.

I'm trying to reach IBM Tech Support,but Tech people are busy..
Any way I'm facing any other issue with ade files.
Randomly jobs are aborting,due to unable to read ade files.
As per my knowledge,RTI Instance automatically created ade files,but some times its not happening.Is their any reason issue...?

I haven't seen any issues with them becoming a problem. As Craig notes, they ought to be cleaned up automatically. Support may be your best avenue here. But let's think about some variables that might impact this.... Anything different about these particular jobs? Server? EE? Are you using IPC stages also? Do you have any of the row buffering options on? The RTI Stages in Server are modified IPC stages, so perhaps there are some interesting issues going on there.........

Ernie