In the previous Industry Insight in this series, I looked at system monitoring and role swapping. This month, I will look at how to evaluate high-availability solution vendors.
Most, if not all, operating systems have features and functionality that you can access and use to customise your own high-availability requirements. But not everybody has the time, inclination or skill required to do this: this is what makes it useful, if not mandatory, to use a pre-configured high-availability solution.
After all, a third-party high-availability vendor focuses all of its attention on writing, testing and perfecting programs and interfaces required to execute and monitor processes and guarantee smooth and reliable operation. As such, they are more likely to give you the outcome you're looking for, while you focus on your business.
Take note
It is helpful at this time to revisit what you should be looking for as you survey the market for high-availability solutions:
* Efficient replication of objects in near-time (as close as possible to real-time). A good high-availability system will also be able to keep a variety of objects replicated in near-time - not just data. You should be able to replicate user profiles, device configurations, spool files and other necessary objects.
* An easy-to-use system monitor that highlights components that are not functioning correctly and objects that are not in synch; it must also automatically reset components and correct out-of-synch conditions. In addition, system monitoring should require no more than an hour's operator attention a day. The word autonomic has particular relevance here: it implies the ability for a system to self-monitor and self-heal.
* An easy-to-execute role-swap process that automates the process of monitoring synchronisation, terminating necessary jobs on the production system and starting all appropriate jobs on the backup system. Here, too, autonomic capabilities are critically important.
In addition, high-availability solution vendors should have referenceable customers who can happily confirm the following:
* That they have performed regular, successful tests of the role-swap, ideally monthly;
* They have had their high-availability solution successfully fail-over to the backup system due to a system failure on the production system;
* They consistently enjoy the requisite level of support from a high-availability vendor. Given the critical nature of high availability, and that most other features are equal or similar, the calibre of the support should be the deciding factor when considering a solution.
In a world of 24-hour operations and growing dependence on data availability, more and more companies are turning to high-availability solutions as they find they cannot tolerate or afford downtime. Even considering the stellar uptime record of the System i (formerly AS/400), which runs at 99.95% around the world, you cannot take a chance with downtime.
Spice of life
In a world of 24-hour operations and growing dependence on data availability, more and more companies are turning to high-availability solutions as they find they cannot tolerate or afford downtime.
Raul Garbini is director of Edgetec
What is of interest to note is that less than 10% of all downtime can be linked to unplanned events, with a mere fraction of that due to a site disaster. The other 90%+ is caused by system maintenance tasks, the kind of thing we've all learnt to live with. This includes data backups, reorganisation of files, vendor and operating software upgrades and data conversions, new application software installations, hardware upgrades and system migrations. If you cannot afford to be down at all, you need a high-availability solution.
When evaluating such a solution, you need to understand its vital components, including how the solution fits into your environment, how much automation is built in, what it will take for your IT staff to manage the solution, and how much time it will consume on a monthly basis.
One last point: a high-availability solution is not the be-and-end-all when it comes to reducing downtime. The size and complexity of your information systems will determine the other factors to consider when trying to reduce your vulnerability to planned and unplanned downtime.
High availability, it goes without saying, is a major part of the overall data recovery/system availability approach, but it can take a variety of software and hardware components to guarantee maximum protection against all exposures and vulnerabilities.
* Raul Garbini is director of Edgetec.
Share