Thursday 1 August 2019

Liberating Legacy Data


Liberating Legacy Data

The latest buzz word in the Industry is Cloud computing and there is a lot of efforts directed towards porting legacy data which maybe residing in the IBM mainframes (banks, financial institutions, automobile companies, insurance agencies, hospital records and even federal government data). Some of us, who have been in the legacy world forever find it difficult to understand why the IBM mainframe, which stood for security, scalable infrastructure and volume computing, is in threat of being made a relic by the new kid on the block-the Cloud. What really happened? The high MIPS cost associated with legacy platforms has always been a pain point for organizations using them. The ever-increasing volume of data requires these investments to multiply with each passing year. Cost effectiveness is the key driver behind the efforts to move data to a new age, more flexible storage solution which can grow as required without burning a hole in the IT budget.

Larger firms have begun experimenting their way out of green screens, some have started moving to cloud, new tech companies are already using cloud in the IaaS, PasS or SaaS models. Few are using cloud together with on-premise to create a hybrid solution. Large financial institutions are venturing to the arena, for e.g. JP Morgan has created a public cloud called Gia. There are still some who prefer UDB or are trying to get data out of IBM mainframe.



As wiki puts it “Cloud computing is the on-demand availability of computer system resources, especially data storage and computing power, without direct active management by the user.” This is where Mainframe gets a blow – with the firms who really want to pay only per usage or increase computing power on high volume days automatically. Mainframe has always been a cash cow – with MIPS usage being billed as per contract and peek time MIPS made even costlier, high volume days MIPS are charged even higher and all firms are bound by a contract – even if a firm manages to cut down mainframe usage we are bound by a contract before its renegotiated. Fast forward to cloud and there is a distinct advantage to the firms – since you pay per usage only and that too at a price significantly lower than Mainframes.

Cloud is not without its short comings. There are major concerns around latency and data privacy which revolve around the following:

  • Most of the cloud service providers are US companies which means that has most of their data centers residing in US and the companies themselves are governed by the US laws. 
  • With the datacenters remotely available in various parts of the world, there is a bit of latency in accessing data from cloud. Most cloud customers aim for data centers to be located locally to reduce this latency.
  • Vulnerability to attacks – since data is transmitted via internet there is a probability that this is subject to information theft. Though advanced security measures are deployed, storing and moving confidential data in cloud has the smallest bit of risk associated with it.
  • Labour and training – Its been said that "Technology is just means to the business”. What the industry will have to do is to step forward to train its current staff on cloud – since business knowledge becomes the core competencies for larger firms.



For a large mainframe of more than 11,000 MIPS, the average annual cost per installed MIPS is about $1,600. Hardware and software accounts for 65 percent of this, or approximately $1,040. Consequently, the determined annual infrastructure cost for a 15,200 MIPS mainframe is approximately $16 million. On the AWS side, using the AWS Simple Monthly Calculator to configure a similar infrastructure to the performance test, we estimated the annual cost to be around $350,000 ($29,000 monthly). This AWS cost could be further optimized with Amazon EC2 Reserved Instances. (Source: aws.amazon.com blogs).





Just kicking my grey cells – is Amazon going to be next IBM since they are already the world leader in cloud computing and management. What IBM did many years ago of selling “Infrastructure-as-a-service” through the release of IBM mainframe is being replayed by cloud service providers like Amazon, Microsoft, Alibaba and Google. With IBM taking over RedHat, is IBM getting back in the game of cloud computing? 

Whatever the outcome - I plan to start writing on a series of tools and methods which can help in the migration of data out of mainframe to cloud or UDB  or Linux. It is time we take this bull by the horn.