IBM Hitches Business Continuity Horse to Managed Backup and Recovery Wagon
December 1, 2008 Dan Burger
The dependence of business on information technology has never been stronger. Planning and investment in one requires the same focus on the other. For some, that link between business and IT may feel a lot like handcuffs. Building a business demands the full focus of some organizations, and although IT may be recognized for providing competitive advantages, it can be a distraction that upper management prefers to turn over to someone else. IBM is happy to help.
Getting the full benefit of IT resources is sometimes better left to others. Business continuity, which includes such things as backup and recovery, disaster recovery, and high availability, is one of those areas where managed services deserves some consideration. IBM’s Business Continuity and Resiliency Services (BCRS) came into existence because there were customers concerned about reducing downtime and the ability to quickly restore their systems after emergency situations. They wanted help. In most cases this means “get this burden off my neck by managing it for me.”
This isn’t a huge number of customers if you compare it to the entire System i installed base. BCRS may be working with 5 percent of System i customers. It could be as high as 10 percent, although that seems unlikely. IBM keeps specific details out of sight.
There are customers that want help getting this set up, but are willing to manage it themselves. And there are also System i users who turn to third-party vendors to address these problems with their own tools and services. Those vendors also estimate their combined market penetration is around 10 percent. Rather than add their 10 percent with the 10 percent that BCRS claims, I’d believe that both sources share approximately 10 percent of the System i customer base.
To get a peek at BCRS, I talked with Debbie Saugen, an IBM System i backup and recovery expert and the self-described technical owner of the backup recovery processes and procedures who provides IBM with customer feedback about what’s technically right and wrong with its products and services. Many of you may have seen and heard Saugen at local user group meetings, on Webcasts, or at technical conference sessions. She’s the face of BCRS. From day one of the AS/400, she’s been involved in the technical side of backup and recovery testing. Her focus now is on the business recovery side.
Here’s what we talked about:
Dan Burger: How does backup and recovery fit into what’s going on in the System i user base as a whole?
Debbie Saugen: Not everybody has a plan. And not everybody who has a plan has a good plan. It depends on how a plan is defined. Is it a recovery plan like “How I’m going to put the data back on the system?” Or is it a recovery plan that includes “Where am I going to go and what am I going to do and how am I going to get a system?”
Most customers do some sort of backup. Whether it is good or not is debatable.
In IBM’s Business Continuity and Resiliency Services, we have a customer base that can come in and test recoveries on a system. They have a contract with us so that if a disaster happens, they can recover on one of our systems. They have a network and they can get their business up and running. I would call that a recovery plan.
DB: How would you describe the customer base? Are they primarily the large organizations with sophisticated IT departments?
DS: Not necessarily. Originally what we had with System i customers was tape-based recovery. And that was all we sold. Then customers were looking for better RPO (recovery point objectives), better RTO (recovery time objectives), and looking to eliminate the problems that are inherent with tape-based recovery. So we have a subset of customers now that do logical replication through our business recovery centers, using software from Vision Solutions. We also do remote backups for them. Some are large customers on a dedicated box, but we have smaller customers that share an LPAR box for that solution. That business is steadily picking up.
These are customers that don’t want to learn the solution or manage it. So we do all of that for them. They may not have the people to devote to it or the technical expertise or both. The center provides the hardware, the network, and the infrastructure that goes along with that.
The difference between large and small customers in this segment is that customers with less than a terabyte of information are the small shops. We have a few of those.
The recovery business has changed from tape-based to logical replication for a variety of reasons. Mostly because companies want better RTO and better RPO, they no longer have a backup window at home so they can do the backup off the replication and they don’t want to deal with shipping tapes any more.
DB: Is there a misconception about eliminating the need for tape backups?
DS: There will always be a need for tape backups. When you start replicating, you only replicate what’s on your home system to the other system. There could be damage. People can delete things. If you don’t have a tape backup, you’ll lose it.
We never eliminate tape backup. We would never recommend that for System i backup.
DB: If there are multiple operating systems involved, how does it effect a backup and recovery plan?
DS: First, I’ll talk about the System i perspective because that’s my focus. Customers are running i5/OS, but on that same box they will have AIX, Linux, and/or integrated Windows servers. We see them coming to the recovery center and recovering that entire environment. If they are running other databases on other platforms, BCRS has solutions for those recoveries, too. They have a plan for the recovery of all the servers. We recover HP and Sun and multiple other platforms.
Sometimes that is why customers are coming to a recovery center because BCRS is not just for the System i, it’s for all the other servers, and it includes the network infrastructure. It’s about having a place for their end users to work. The System i is only a portion of our BCRS business. A disaster recovery plan involves all the servers a customer has, all of the network infrastructure, getting the entire environment you have up and running. But you still have to narrow it down by platform and how each one will be accomplished. This is rarely a standalone System i project. There are some customers that only have a System i, but not many.
DB: How are decisions being made about the extent of replication that is being done?
DS: When I talk about replicating data, I’m talking about replicating applications, programs, data, etc. We can replicate everything. The organizations look at how much they want to invest in a disaster recovery plan and that’s based on the business impact of downtime. Most times, there is a business impact analysis that looks at how much money the business would lose if the systems go down. The executive team weighs the cost of what it takes to keep their business up and running.
Part of this has to do with how quickly you recover. The other part has to do with how much is being backed up.
When you do the business impact analysis, you determine what you absolutely have to have to keep your business up and running. For instance, some won’t include development in a disaster recovery plan. It will be a loss if the development server is down, but it doesn’t impact the business as much as production.
How these things are viewed changes from industry to industry, but everyone makes decisions based on what’s critical to the business.
DB: Is there growth in this backup and recovery area? It seems that industry insiders have been talking about it being between 5 and 10 percent of the System i market for years. It’s hard to see progress based on those numbers.
DS: I think it may be above 10 percent. I say that because there are more people doing this internally–managing it themselves.
I work with a lot of customers that are dealing with the tape-based recovery and the issues and the time involved. They are looking for something better. I lead them to logical replication, because it gives them the best RPO and RTO and it gives them the capability of doing remote tape backup.
DB: And what about the cost?
DS: Having a full disaster recovery plan is expensive, though. It will cost at least three to four times more than current tape backup.
From my perspective most customers want a full-blown solution. If you are going to do backup, you want disaster recovery, too.
With BCRS, the system is up and running in a new location as soon as the switchover is completed. Now you have all the data up to the last replicated transaction. It’s more expensive, but it works very well.
Midrange Shops Not As Protected from Disaster As They Think, Vision Finds
Midrange Shops Get Disaster Recovery Services from IBM
The System i High Availability Roadmap
You Have Life Jackets, But Have You Ever Put One On?
IBM White Paper: Remote Backup Alternatives for System i