DSXchange

The Datastage job is having SFTP Stage, connects to Mainframe to extract data from mainframe to Linux. We are using TECTIA to transfer the data.

There are many jobs which does the same functionality but everyday there will be few failures and after re-start couple of times , jobs run successfully.

Error message

Server> INFO (49479) SFTP_MVS,0: error in readCommFifo or ftp command function
2018-11-16-01.49.22
Server> INFO (49480) SFTP_MVS,0: Output 0 produced 0 records.
Server> WARNING (49481) SFTP_MVS,0: subprocess failed with exit code 5,SFTP_MVS1
Server>,FATAL,(49482) SFTP_MVS,0: The runLocally() of the operator failed

The jobs were running without issues from past 11 months and started to fail from past 1 month. There are no changes/upgrades in MAINFRAME OR LINUX SERVER for past month. Appreciate your help if anyone see similar kind of issue and resolved.

Thanks

Well... whenever I see any post like this where things mostly run fine but occasionally fail but then work when rerun, that says "resource issue" to me. Now, don't ask me what resource here, but I'll wager there's possibly a kernel parameter or something of that nature that needs to be bumped up. And it doesn't necessarily mean something changed other than perhaps the workload... the total workload... on the server.

What Craig said, and for us it was the active setting on the FTP host server. It limits the number of available data ports, and as volume and usage increased here, we started seeing a gradual increase in abends on no ports available. Often, it was masked by the host deciding that the file wasn't found. It took a while to figure out the real cause.