Today mainframe users are faced with a number of challenges such as the increasing post-warranty costs of maintaining their tape infrastructure as tape silos approach end of life or are out of warranty, the need to reduce power consumption in the data centre, struggling to match the ever growing amounts of data with a robust backup infrastructure, failing to test DR policies frequently and regularly, and of course running out of floor space or finding the cost of additional floor space prohibitively expensive.

As a result organisations are looking to eliminate their dependence on physical tape and on off-site vaulting of tape, especially since governance requirements have enforced stricter adherence to data management and off-site storage regulations. Today mainframe users are demanding faster recovery times even though business practices and data growth have lead to shorter backup windows and yet backups are taking longer.

A survey, carried out in September, yielded the following answers:

  • Approximately 50% of the respondents had an IBM mainframe tape silo; Oracle (Sun/StorageTek) accounted for the other half;
  • 36% of these tape silos were reaching end of life within six months, becoming a major management and cost issue going in 2011;
  • 73% of organisations also used some form of virtual tape;
  • 91% claimed to test their Disaster Recovery (DR) environments at least yearly but;
  • 9% said they do not test at all mainly because the complexities of recovering tape-based data often limit or preclude proper DR testing and due to the issue that remotely recovering from tape can be slow and awkward, particularly if recalling data from HSM ML2 tapes;
  • 55% of mainframe storage administrators would like to improve their HSM migration and recall performance, replacing a mechanical process with an automated tiering underpinned by the latest solid-state technologies because by switching away from their old tape silos to disk-based de-duplication they will be working at electronic rather than mechanical speeds;
  • 64% had floor space constraints;
  • 18% of respondents were keen to reduce their offline tape transportation (white van) costs and CO2 emissions.

At present mainframe users are facing many of the challenges encountered in other IT environments, such as continuous and significant data growth, the pressure to backup, and the cost of staff, energy and floor space.

Because mainframes have been around for longer than distributed computing, when disk was much more expensive than it is today, tape has been the traditional backup medium of choice for their administrators. However, switching from tape to disk-based de-duplication today can save end users up to 90% in floor space in a traditional silo environment.

In addition, moving to a disk strategy makes DR testing much easier and quicker, increasing the reliability of data protection plans and helping businesses manage any disasters more effectively.

There are more than 10,000 organisations around the world still relying on mainframes today, and most of these users are not planning to move away from their mainframes because of its much higher reliability, security and scalability than that offered by distributed systems. And yet we are seeing few other vendors bringing out new solutions aimed at the mainframe space. It is truly puzzling because we have had significant success in addressing this market.