Living in a time of war, weather changes, and presidential primaries is not for the faint of heart. The question is not whether your family and house will still be there when you get off work at night. The question is whether your office will be there when you show up for work in the morning!
I have written previously about the effects on companies of the 1994 California Northridge earthquake and how the systems administrator of one area pet food company saved her firm by slipping a backup tape from the company's AS/400 into her purse and taking it home with her the night before the earthquake. The stories of companies that have been wiped out by fire or flood-and lately hurricanes-are legion. Let's face it. These days it seems like it's not a question of if something is going to happen but more a question of when! Who can forget the chilling images of the Twin Towers coming down on 9/11?
Companies that have highly critical data-usually financial services firms and healthcare industry providers-are extremely worried about losing data. Unlike most of us, they are not concerned about losing just a disk drive. They are worried about losing the entire site where the data is stored-building and all. It could easily happen in an earthquake or tornado. And floods don't do your computers much good either. What to do about it?
High availability solutions abound on all platforms. This probably is because most platforms traditionally have been a little less than reliable. In the System i world, where the computer's claim to fame is its reliability, stability, and endurance, what's to worry about? Well, there is plenty to worry about, actually.
In the early '90s, IBM introduced remote journaling into the AS/400 operating system. The only problem was it was a bit difficult to configure and use. Then, in stepped the high availability solutions providers Lakeview, Data Mirror, and Vision Solutions. iTera and Maximum Availability formed later as challengers to the big three, offering limited solutions at lower prices. They took the IBM technology in the operating system and made it simpler to employ. Transactions that were input into one system were almost immediately "mirrored" onto a second system, which could be either at the same location or at a remote location.
Several years later, IBM introduced clustering into the midrange market. To put it mildly, it didn't catch on; it was ahead of its time, at least in the AS/400 market. It took a maturation of the AS/400-morphing into the iSeries and then the System i, with faster processors and more storage and networking enhancements-for clustering to make economic sense.
Today, Vision Solutions is finding a market for three-node clustering solutions that may indicate an emerging desire for increased security and near fool-proof disaster recovery solutions that is truly a sign of the times.
Just this week, the company announced it has been collaborating with IBM on System i cluster solutions for customers in the U.S. and abroad. Prescription Solutions, a leader in the pharmacy benefits management industry, decided to entrust Vision Solutions with its vital systems in a three-node topology disaster recovery cluster solution. Senior Systems Analysts Ron Yohe and Reid Parker of Prescription Solutions say that the Vision clustering solution in MIMIX "has greatly enhanced and simplified the switch process for our three-node installation."
In an interview with MC Press Online, Craig Johnson, Vision vice president of research and development, and Bill Hammond, director of product marketing, said the three-node disaster recovery topology makes sense for customers who want an extra measure of protection.
"Clustering is still in fairly early adoption from our customer perspective," noted Johnson. "But it's been our experience this [past] year that the adoption curve is beginning to pick up. More and more customers, particularly at the high end, are deploying these three-node type environments."
Companies that have critical data and can afford three nodes are starting to think about using the technology inherent in today's System i hardware and operating system, he said. "For our accounts in this three-plus node environment-more complexity, higher needs from a switching perspective...clustering makes sense."
Johnson says the desire to have a truly fault-tolerant system is driving System i sales as well. "The reason they are buying that third site is they want another level of redundancy, or fault tolerance," he said. "In the event they have to take one system offline, whether for upgrades or an unplanned outage, they still have another level of availability when they go to that backup server...they always have a level of redundancy that they wouldn't have in a two-system environment."
One would think that with two backup machines there would be little need for backing up with tape. Not so, according to Hammond, who notes that "the archival backup medium is still strong, and people still use that... I think the replication of data in real time-which is what we're able to provide-does allow some flexibility when it comes to recovery...but we don't see a lot of our customers completely getting away from tape backup."
Regardless of IBM's acquisition of high availability solution provider Data Mirror and continued high availability enhancements to the i5/OS operating system, Vision says its future in the area of HA and DR is brighter than ever. As IBM continues to introduce more high-availability technology into the midrange platform, users' appetites for straightforward solutions and easy management of it will increase.
And with the direction the world is headed, one wonders if even a three-node fault tolerant system is safe enough to guarantee business continuation. There are just too many kooks out there.