Unlike Disaster Recovery, High Availability is all about keeping the system available to users, and nowadays that means internal users as well as external users. Historically, high availability solutions could only be afforded by large enterprises; e.g. banks, but advances in remote journalling (from OS/400 Version 4 Release 5 onwards) mean that high availability is now easier to use and less expensive to manage.

Disaster Recovery is discussed elsewhere, but is very much reactive to a problem, and recovering hardware is only part of the recovery process. High Availability takes care of both loss of data and unscheduled downtime, so if your business can survive loss of systems for 24 hours this remains a viable solution, but if unplanned outages mean loss of business then you need to be considering a High Availability solution, one which reduces system downtime from days to hours, or even minutes.

Availability is usually expressed as a percentage of "uptime", and worthy of consideration is that 99.999% availability means having a maximum systems downtime of 525 minutes, roughly equivalent to a working day, over a year.

The following table describes a number of approaches, all based upon MIMIX ha Lite© software, from Vision Solutions, Inc. This proven software is used thousands of businesses worldwide, and exploits the advanced journaling technologies within i5/OS (OS/400) by efficiently replicating changes to the backup server in real time.

Emerging technologies such as Vaulting and Continuous Data Protection are also outlined.

Product Highlights:

Maximise system availability across an entire organisation or supply chain so that companies in different time zones, customers or suppliers have access to company data 24x7x365. Immediate switchover between two processors allows scheduled downtime for upgrades and other routine tasks.

Provide Disaster Recovery solution at a fraction of cost of traditional methods. Make use of the old iSeries processor, using it as a hot standby (DR) processor. Here we are not talking about taking a handful of tapes to some remote office location, adding the time taken to reload, and then paying accommodation charges and expenses. It's all expense and nothing to show for it unless, of course, disaster strikes.

Isolate information provided to a web server from attack. Key information can be replicated in real time from the primary iSeries server to a web server, thus protecting primary data from potential external threats.

Vaulting is a technique whereby the backup process is fully automated, with data being transmitted to an iSeries processor located at CVSIs premises. The business proposition is that this approach provides more system uptime, requires no user intervention, no additional hardware, and no additional staff resources.

This approach will also be of interest to companies whose regulatory policies demand having off-site backups.

The business value achieved by replicating data is the bedrock of high availability solutions, but what happens if there's a corruption or deletion on the master processor, which then gets replicated to the HA server?

Despite best efforts to plan against unforeseen contingencies, the chart reveals that there's still a 26% possibility that data can be lost through human error, whether unintentional or otherwise.

Under these circumstances, recovery of data can be achieved using Continuous Data Protection (CDP), capturing every change, and allowing users to recover to any point in time.

The Vision Solutions offering (MiMiX CDP) currently supports an Optimized Recovery Window of 60 days, which is probably beyond anybody's realistic expectation, or disk capacity. It is free with MiMiX ha1 and MiMiX ha Lite.

Please contact us here to discuss your High Availability requirements.