Replatforming

Replatforming is the activity of making an application currently installed and running on a specific operating system available for another operating system. As an example, let’s take an application that runs on AIX now and plans to run it on LINUX in the future.

Why is this required?

There are several reasons which can also occur in combination

• the original required hardware is no more available or becomes too expensive
• the cost for the operating system, the license cost for middle-ware or related maintenance contract costs are increasing
• the performance of the existing system is no more sufficient and a vertical scaling is not possible or too expensive on this type of hardware
• the hardware and application software should be outsourced to the Cloud and there typically LINUX based container systems are in use

Which options do we have?

On the hardware side the concept of virtual machines is helping as long the original operating system (in my example AIX) would be available also as guest operating system. This would solve the hardware dependency, but not license costs to be paid for the original operating system.
On the software side, it would be very beneficial and would give a great flexibility if the used application software is directly available for the new platform with the same functionality. The last point is important because if not all features are available then we have no 100% replacement. Software applications which are developed on different platforms in using different programming languages are practically impossible to keep identically on their functionality. We all know that even software which was working good in a certain version could behave a bit different in the next version.

Replatforming with Papyrus

The Papyrus software portfolio has been designed and developed for 35 years in such a way that we can offer back-end software for every common operating system. Today we talk about

• Windows
• Linux – Red Hat, Suse, IBM z Systems
• Unix – AIX, SUN
• z/Os native for some products

We have made sure that we not only have the same interfaces to the outside world, but that we also use the same source code for all supported platforms. Only the core layer, which communicates directly with the operating system, has been adjusted accordingly. This is possible by using the programming language c/c++ and there are compilers on all platforms.

Not only is the source code the same, the file formats created are also identical for all platforms. For example, the Object Spaces of the Papyrus Communication Platform can simply be copied from one operating system to another, e.g. from AIX to Windows or LINUX, and the Papyrus Node running there can be started directly with it.

The Papyrus Platform has the flexibility that the participating nodes can run with all service tools on any operating system and also in any mix. This is very useful when a specific hardware device driver (a printer, a scanner) is only available for a specific operating system, or when a third-party application is only available for a specific operating system. In this case, only that particular Papyrus Node can stay there, the rest can be moved. Hybrid systems that mix on-premises with the Cloud are also possible.

What is the effort for a switch of the operating system platform?

Let’s make it short, not much.

The time someone needs to install the Papyrus software on the other platform, and the time to copy the Object Spaces from the old system to the new one. Even the product authorization can be moved without any additional cost within the maintenance contract.

Michael Firnthrat

Development Manager at Papyrus Software