In chapter 1
, I described how the planet is becoming ‘smarter’ and that this transformation is creating enormous amounts of new data that needs to be effectively managed. In this chapter, I’ll review some of the things that complicate the effort to ensure all this data is properly retained, protected and available when needed.
Ideally, you would like to have a single tool that does everything, across the entire enterprise, providing the ability to effectively respond following any type of event. While many vendors promise to solve your problems, nobody can provide this capability in a single package – the problem is just way too complex. But (tease), IBM is driving toward a unified recovery management capability that enables you to manage a selection of tools from a single administrative interface. More on this next week; first we need to ensure that we appreciate the complexity.
The first category is infrastructure – where is the data?
Your IT shop probably includes several if not many types of hardware: computer platforms such as x86, Power, RISC, Sparc, mainframes, etc. And there are a wide array of storage platforms, including direct-attached (DAS), network-attached (NAS), tape, and I’m sure many of you still have optical disks somewhere. And from many different vendors!
On these platforms, you’re going to have different operating systems: AIX, HP-UX, Linux, Solaris, VMware, Windows, z/OS. Then they’re going to be physically located in different places – data centers, staff offices, production facilities, remote/branch offices, disaster recovery sites, and warehouses.
Different types of networks, and the available bandwidth on them, further complicates the system. You have local-area (LAN), wide-area (WAN), storage-area (SAN) and metro-area (MAN) networks; additionally you may have cable networks running to some offices (particularly home offices), telecommunications networks that now carry data, and USB connections to some storage devices. And finally, you likely have important data being created and stored on user workstations.
How many tools do you use just to cover this level of complexity? But wait – there’s more! The next question is: who owns the data?
We also have to matrix in the different type of applications you have – general file systems; email, instant messaging and collaboration systems; databases such as B2, Oracle, SAP, SWL and mySQL; and your industry-specific mission-critical applications such as CAD/CAM, medical records management, software development, manufacturing resource planning (MRP), or customer relationship management (CRM).
Now consider that the data created and used by any of these applications may be on any hardware platform and operating system, in multiple locations, using a variety of networks. A lot of the data may be on user workstations as well. Oh my!
But there’s still more – what can go wrong?
As I noted in my last blog, lots of things can go wrong, any you really need a different response for each of them:
- Lost files or e-mail. This may be inadvertent or malicious, but it happens a lot. You need a way to quickly restore individual data objects from a local backup copy; otherwise you’ll be spending a lot of time and money meeting user expectations.
- Database corruption or virus attack. These can go undetected for weeks, so you need a way to go back in time, and restore the data to a point prior to the initial problem.
- Hardware failure: It’s a fact – disks and computers sometimes break; a fast volume-level or system-level restore capability is needed to limit the downtime of critical applications and to ensure your continued employment.
- Local disaster: Fires, floods, power failures do happen occasionally; you need a copy of your data in another location, and you may need the ability to quickly restart operations in another office.
- Regional disasters: Far less common, but massive earthquakes, power grid failures and wars can disrupt business in a number of offices across a region; multi-national organizations should have contingency plans for even the worst-possible situations.
OK, now draw a line from every block on the diagram above to every other block and tell me what your backup and recovery plan is for every line – even in this simple diagram, there are 100 different scenarios, but when you consider all the variables, it may be millions. What tools would you use, who will use them, how long will it take to recover, and how much data will be lost? And what does it cost?
More on that next time!"The postings on this site are my own and don't necessarily represent IBM's positions, strategies or opinions."