The workload challenge

Looking to bin the mainframe

The company I work for (a global telco monster) has a mainframe for all our provisioning / order entry / billing / maint. etc and it works just fine.

Over the last 5 years they have spent many millions on a new web based system, you know the sort of thing, point, click and it's designed the circuit for you. In fact what they want is for the order entry people to put the info in, the system to design the circuit and then apply any connections etc that may be required.

That is feasible if you were starting a new company from scratch with an empty datatbase but if you import the old data from the mainframe (which in some cases is wildly inaccurate) then you'd have to shut the company down for 6 months (at the very least), do a complete audit of every single bit of kit, and even then we have so much legacy stuff (ancient kit) that may or may not be able to use the command language they have chosen.

The new system doesnt work mainly because the people designing it don't understand telecommunications.

The old system has it's problems of course but EVERYONE in the company knows these limitations and the work arounds to them.

I actually want them to keep the mainframe, it works, it is more than adequate and everyone in the company knows how to use it.

The company wants to shut down the mainframe at the end of April and use the new system (No, it still doesnt work properly and no, there has not been one sniff of training)

I'm not a mainframe / IT guru type but I'm sure I have read about alternatives for replacing a mainframe with blades and running the same stuff on it, likewise for running fancy front ends to make the mainframe more accessible. I'm sure either (or both) would be preferable to the $hit storm about to hit us in the next couple of months!

Anon for obvious reasons :-)