As technologists, we are not mainframe fanatic actually.   We always face products platforms as objects of technologies, so they are equals in our point of view.  Nevertheless, we must be realistic that our main asset is our skills in mainframe technologies, especially in system-level programming which is an expensive asset not easy to have it.   Besides, based on our and many other survey results, mainframe still dominates in mission-critical computing applications such as banking and other finance services systems as well as public services and governmental information systems    

Especially in large banking industries, not only in mature countries, even in Indonesia, the number of mainframe sites still higher than any other platform.   Currently there are at least 4 mainframe sites in 4 large banks in Indonesia.   Compared to the number of each individual non mainframe platform, mainframe is still dominant.  

Moreover, chances of system-level software business according to our way of thinking relatively better in large-scale computing systems as in fact not many people are playing there.  Especially for mainframe, the most popular large-scale computing platform, meanwhile, skills around that is really shortage.   All of those reasons convince us that mainframe is the best choice for us to be placed as our main software industry role.   

Although Deru (NSI founder) has quite reputations in digging the stomach of mainframe OS since 1986, even ever built a “miniature” of OS for IBM System/390 architecture and written 3 tight books discussing about assembler, mainframe computer architecture and systems programming, it was not very easy to convince the team to setup a real “system-level” software industry.   A real system-level software product development project needs more than just high quality skills.  It also needs highly keenness and fitness for everyone to finish quite long term project with several difficulties since system-level software deeply touch quite detail OS and hardware architecture, instead of just its main application logic.  Sound makes sense, isn’t it?    

Deru, therefore, has to prove it himself.  He then used his client’s challenge by the time to improve the existing scheduling system to support job-step level triggering.  This was very good challenge that scheduler and automation are mission-critical areas where not every mature vendor quite confident to play in.  He then went to rural area in Pemalang, Jawa Tengah to find a good place for meditation inspiring himself. Finally he found Semingkir, a small beautiful county in a valley near to mount Slamet in Pemalang.   He created side activity first to hide his main mission from people around there, by developing distiller machine and establishing patcholli oil distillation business there.   Soonest that business ran, he started his main mission, designing and coding the scheduler and automation tools from his laptop.   He did it alone.   

Fresh air and beautiful views of Semingkir, help him tuning up his capabilities in digging the traffic mechanism of system events beneath the surface of z/OS to find the best algorithm to manage them simultaneously.   In certain situation when met difficulties, he drove his Jeep into the jungle around the valley to get better inspiration from the sky.   Offroad or site seeing sometime much better than sleep to get the brain refreshed.   

Two years later, 2 products in a single framework were ready for testing, and he called Sekar for automation (events manager) and Puspa for scheduler, both are his daughters’ names, and the whole package was then registered as his intellectual property with name XDI2, later is known as zJOS-XDI.   Thanks God, he has proven himself not just to finish the project, moreover, yielding the product for mission-critical computing system, which certainly needs very high reliability.   He then called Semingkir as his “Silicon Valley”.       .  

System-level software is software which works tightly with internal layers of operating system (OS) and/or hardware, for example, event-driven automation and scheduler tools, security tools, systems monitor tools, database engines, data networking carrier tools etc.   Besides its type and style of application, it needs deeper systems programming skills to develop, whereas less people have it.  It more sensitive to the system as it needs to run in privileged mode when accessing internal system resources.  If it get problem, it can potentially impact to the whole system.  So why is interested? 

Unlike user-level software, such as banking, insurance, payroll, accounting applications etc, which in fact easier to develop, more manageable risks and easier to find skillful people to involve.   Technical difficulties and potential risks of system-level software are totally not interested in common point of view.   Hard to develop, hard to support, hard to find the skillful people to involve and have to tightly watch day-to-day hardware and OS changes.  

In today’s business point of view, system-level software, however, is more interested.   Not many people around in this world, so lesser competitors.   Hard to find skillful people, so not easy to start such business, although it’s more interested today.   Luckily NSI was founded by such person!   So it can start by him alone.   And surprisingly it came up with automation and scheduler tools products which have been proven work stably in the mission-critical areas.   

zJOS-XDI© is an integrated automation software package for z/OS platform.   It provides 3 major functions, (1) systems events manager, (2) events-driven workloads scheduler and (3) spool/output distribution.   Each major function is marketed as a separate product. 

zJOS/Sekar - systems events manager, to manage users' defined automatic action(s) against occurrences a certain case of a type of system-event.   Sekar recognizes messages, commands, job-step termination (batch job or STC) and job termination (batch job or STC) events, and date/time schedule.    

zJOS/Puspa - events-driven workloads scheduler, to manage users' defined automatic schedule(s) based on occurrences of combination of a number of cases of types of system-events.  Puspa recognizes messages, commands, job-step termination (batch job or STC), job termination (batch job or STC) and dataset access (open) and dataset release (close) events, and date/time schedule.   

XDI/AutoXfer - spool/outputs distribution, to manage users' defined spool/outputs download and/or distribution based on their occurrences in selected classes and/or with selected writer name and/or from selected job names.     

zJOS-XDI is featured with integration facility.  In networked multiple z/OS hosts, you can automate all events managements and scheduling systems integrally at a single control with agent technology.  Occurrence of an event in any host can trigger any set of actions or schedule in any or several hosts.   Schedule of a job can be based on multiple occurrences of events in any or several hosts.   

As solutions to mission-critical application environment, zJOS-XDI has been proven in the real production environments in Indonesia.   It’s very user friendly and doesn’t require special skills to operate.   No scripting is needed.   Replacing your existing mainframe automation tools with zJOS-XDI means modernize your system as well as extremely cut the costs.