1:53 AM, and Sacha's phone was ringing loudly. The woman on the other end of the line was trying to speak calmly, but the fear in her voice was obvious. It was clear that this was the first big production failure she'd had to deal with.

"Production job SYSSYNC_VMX_PROD02 failed, it says something about a failure with... JCL? Er, was it TSS?" Embarrassed and flustered, Leah confessed that she forgot to bring her notebook with her and couldn't remember the specific error.

Sacha tried to be reassuring. "OK, good, so it's nothing to worry about." Oh god oh godohgod, he thought. "At least it wasn't SYSSYNC_VMX_PROD01 that failed!" What the hell is SYSSYNC_VMX_PROD02? Is there even a PROD01 job? "We have a six hour SLA, which should give us plenty of time to fix it." God help us if we can't fix it in six hours. Sacha wondered if he was failing at sounding calm as much as Leah was.

SLA Compliance

There's failure to comply with SLA at your job, and then there's failure to comply with SLA at a military job. You're probably thinking "no, the consequences are pretty dire if someone screws up at my job." Yeah, cry me a freaking river. You probably have to fill out some paperwork, talk to the bosses, and get on with your life. In the military, you slip up, and R. Lee Ermey screams at you until you're ready to cry, then your coworkers hold you down to your bed and beat you with pillowcases full of soap. Then you have to fill out some paperwork and talk to the bosses. Leah was right to be worried, as was Sacha.

Despite having heard and occasionally using all kinds of arcane acronyms, Sacha didn't know what JCL, TSS, TSO, or CICS were at the time. Fortunately, he had the on-call mainframe guru at his disposal. After some back-and-forth, the mainframe guru reported that it was just a faulty tape mount issue. "All you'll have to do is change this parameter in the JCL file and re-submit the job for execution." And so Sacha did. And the job hummed along nicely. He called Leah, who was clearly relieved at the good news; crisis averted with hours to spare! At 3:10, he hung up his phone, happy to return to sleep, hoping to have the dream where he's an eskimo.

A Late-Night Workout

Three and a half hours later, Sacha was living it up in his dream igloo when Leah called back, panting heavily. He briefly contemplated screaming "pervert!" and hanging up, but decided against it. "What's up, Leah?"

"Mainframe *pant* robot *pant* constantly *wheeze* asking *pant* tape *pant* replacement..."

"Whoah, slow down," Sacha said. "Did you engage in mortal combat with the mainframe robot?"

Leah was clearly unamused, but was still more concerned with catching her breath.

Sacha learned that since the "quick fix," the library had exhausted its internal mount capacity. "It asked for a tape *pant* after the JCL file change *pant*. I went to the *pant* supply closet and grabbed one *pant*."

"The 204-B closet?"

"Yes *pant*." As she described the situation, she was able to talk for longer stretches without as many deep breaths. After she'd grabbed the first tape, it didn't take long before the system asked for another tape, sending Leah back to the closet, and then back to the machine. This process kept repeating itself, with the system launching thousands of tape requests, munching tapes like popcorn.

The poor girl had run a 30 meter dash some 400 times to keep the system fed with tapes. 400 times running 60m (30m each way) was almost 25km. Leah was still breathing heavily after describing all of this; she sounded like she was ready to collapse at any moment.

Sacha called the mainframe guru again, and the three of them worked out a different solution that thankfully meant less running for Leah. But on the bright side, they maintained SLA compliance, and Leah didn't have to go to the gym that day.

[Advertisement] BuildMaster allows you to create a self-service release management platform that allows different teams to manage their applications. Explore how!