IBM FTP Command using *SSL

We use the IBM supplied FTP command to do a lot of file transfers to windows servers. Many of these file transfers run automated late night or early morning when we are not open. Recently we started seeing dumps from these jobs. A QPDSPJOB and a QPSRVDMP spool file. Error issued is C2M1601 Signal SIGABRT raised (abnormal termination). We would run these same jobs when we start in the morning if we see them fail overnight and they will run perfectly fine in the morning. We have been trying many different things to try and isolate this issue and thought for sure it had something to do with the windows server being the FTP Server in these cases. At times an FTP download would fail and an FTP upload from the same location on the windows server would work. We would even try watching the jobs from home at night and when one failed we would manually run the FTP command from a command line and it would fail with the same error and 10 minutes late it would work without any changes made anywhere on either side of the process. When we could see the failure if we changed to a non-SSL connection it would work every time. (FTP RMTSYS() SECCNN(*SSL) vs FTP RMTSYS())

Just today for the first time it failed on an FTP I was doing from an interactive 5250 session. Tired it three times from one session and it failed three times with the error we see in the automated jobs. I used a second session I happen to have started and it work correctly immediately. Used the IBM FTP command from the session where it worked and that worked too both SSL and nonSSL. Went back to the other session where it didn’t work and IBM FTP command failed also on SSL bot not on nonSSL. Signed off the failing session and back on and tried FTP again and both failed while the other session worked fine in all instances. Ten minutes later (the time I was writing this email) the session that was failing is not working and I did nothing.

This almost seems like something with an SSL connection is being cached outside the interactive session that eventually gets cleaned up and then it works again. If there were an incoming HTTP request I would suspect one of the service jobs in the HTTP subsystem was getting confused. Is there some autostart or prestart service job that all FTP traffic might use/reuse in QSYSWRK that might be where our issue exists?

This one really has us stumped.

Mike Cunningham

p.s. search for the C2M1601 error didn’t find anything related to FTP on a V7.1

________________________________
This email may contain confidential information about a Pennsylvania College of Technology student. It is intended solely for the use of the recipient. This email may contain information that is considered an “educational record” subject to the protections of the Family Educational Rights and Privacy Act Regulations. The regulations may be found at 34 C.F.R. Part 99 for your reference. The recipient may only use or disclose the information in accordance with the requirements of the Federal Educational Rights and Privacy Act Regulations. If you have received this transmission in error, please notify the sender immediately and permanently delete the email.