DSXchange

There is a section in the IBM Knowledge Center web site about manually removing the software from each tier. Follow all those steps then reboot. After removing software from a Windows machine, I always like to run the CCleaner utility to clean up the registry. It scans the registry and often finds h...

This page has the error, cause, and resolution (not sure if you're using this particular method of connection or not):

http://knowledgebase.progress.com/articles/Article/6058

Go for a Server job.

I usually just look at the user name in the detailed job log entries, although what shows up there can vary depending on user registry type and whether or not credential mapping is used.

Best case: Get your own ID with access to run the "sudo su -" command then you are free to do all the pre-req steps, install, stop and start, and maintain the application ongoing. The bottom line is that someone who knows the software very well, and has time allocated to properly support t...

Every place I have been has not allowed a direct login as root due to security. You always have to login with any other ID that has been given sudo access to switch user to the root user, then you run through the necessary commands and pre-checks, like umask, ulimit, ISA Lite, etc., then run the set...

The 8 minute delay between job log entries is just a possible misinterpretation. The detailed job log entries show up in the log ... eventually, when they feel like it. It does not mean that 8 minutes went by between a node 0 insert and a node 1 insert. I often see this situation where one node's SQ...

I have definitely run into this issue before, due to special characters in the password. Thanks for sharing.

Yes it just leaves a little too much up to my imagination. :wink: A really, really bad call in real life could go on for 50 hours and 30 seconds. Just curious in the data if that would look like 3000.30 or if calls > 59 minutes may be filtered out. If the latter, then the conversion logic is simpler.

This looks like a Server job question; StringToTime is a parallel function.

So just to confirm, what does the actual input data look like for call times over an hour? That example makes a difference in your logic.

If you have already pieced together a solution, congratulations! The Big Data File stage allows one to generate multiple files and specify labeling and processing order (per the documentation). I have not had opportunity to test that yet. I do not think it could be based on data in one of your field...

I would never consider that to be normal but I have seen the same behavior from time to time. I always kill those "runaway" processes. Also as ArndW mentioned, kill client processes that are older than a day. We request our developers to close their clients each day before they leave work....

Could you take an intermediate step to loop through the list of files and move or copy each one into an empty subdirectory? Then your job can point to that location using a file pattern to read *.txt.

bicap wrote:This is comparison of current record with previous group and next group.

If you can think of it in terms of a comparison of the current record with the previous group only, it may simplify the logic.

For your information: The APT_IMPORT_HANDLE_SHORT environment variable appears to be new in version 11.3.x. If you set it when reading a file, it is supposed to assign default values, or assign null if nullable, to columns that are missing, instead of rejecting such records. More details are availab...