Companies who operate IBM’s System i platform know they have one of the most reliable hardware platforms available. Reliability, however, is not the same as availability.
In the past, high-availability (HA) solutions have been sold primarily to large enterprise users. Although enterprises remain a significant group, there’s been a substantial growth in high-availability adoption in recent years from small and medium-size businesses (SMB). The needs of enterprises are different from that of SMBs however, and therefore optimised solutions are required for each of these two different groups.
Large enterprises tend to have sophisticated environments and in-house technical resources to manage these environments. They need high-availability solutions that have maximum flexibility and performance. These solutions have been built from ground up for the needs of enterprise users.
SMB customers tend to have less complex environments and have fewer in-house technical resources. They want high-availability products with self-managing capabilities.
Also, business change is an important factor. Companies restructure quite often and it’s important to many companies to have a flexible high-availability solution that can easily adapt to these business changes. High availability is a business solution, not just a technical solution, and offers significant business advantages and strong return on investment.
High-availability software is an element in compliance with government regulations in many industries – financial services, telecommunications, and healthcare, just to name three – and it is increasingly important in a virtualised server environment, too. One of the benefits of using virtual servers instead of physical ones is that workloads can be moved around not only inside of a machine, but also across physical machines that have shared storage.
In reality this means that companies can optimise their applications, data and workloads on fewer physical servers. The cost savings of such server consolidations can be significant. It, however, creates an “all eggs in one basket” environment, which increases the need for – and importance of – a solid high-availability solution.
A high-availability system is made up of two or more machines (the primary and one or more backup machines). Each machine is considered a node in the cluster. Such high-availability configurations provide shutdown and automatic restart of unresponsive (failed) software programs, and migration to the cluster backup node when failures on the active node are detected. The cluster backup node assumes the cluster name and IP address and automatically takes over system processing until such time as the failure is corrected.
Some high-availability solutions require more hardware and communications bandwidth, and this can be quite expensive for many customers. For example, with some solutions you can do tape backups without taking users off the system first to get a reliable checkpoint. With other solutions, you cannot. You can perform recovery tests while users are online, you can do workload balancing, you can place the recovery server any distance from the production server without impact on response times, and you don’t have to restructure your environment. But only with certain configurations and solutions.
It is critical to consider data consistency, and what it means to your business. System i applications were designed around the concept of business machines processing business transactions on single level storage. Whereas some high-availability solutions mirror disk sectors, a more effective solution mirrors business transactions. Business applications don’t recognise disk sectors so you will first need to repair data consistency otherwise the applications won’t work, which could take a lot of time or in some cases it may not be possible at all.
For example, if you had an unplanned system failure during end of day processing you typically want to restart the processing on the backup system from a known data consistency point. That’s easy to do with the right solution, but not with others because you need to regularly interrupt production users and make copies of databases.
The high-availability landscape is very dynamic and very vibrant. There are hardware solutions that come into play, and there are a number of players who address the market opportunity with a lot of innovation. The key is choosing a high-availability solution that is right for you, easy to use and transparent to end users.