Great blog post from BernieT from Datacore…
High Availability – Where is your weakest link
I find the primary focus for investment is in Server hardware and software, utilizing server virtualization, clustering and image based backup solutions to provide redundancy/recover-ability for business applications. Many area’s are overlooked and ofter become the Achilles’ heel of IT environments.
Let talk about centralized storage;
While many are quick to invest in multiple servers to build redundancy and “spread the risk” in their IT infrastructure they are complacent when it come to placing their critical data (which without, your servers have no use) on a “single box” array. Yes they are highly available with “internal redundancy” but are a single point of failure and are susceptible to outages, planed or unplanned more often that not, caused by “environment” issues such as power, cooling, air quality, water, building.
The next logical focus is “Disaster Recovery” capabilities, however there is a great deal of complexity and cost associated with having to plan, build, and maintain a business continuity plan, it involves applications, data, hardware, communications, key personnel, facilities.
Doesn’t it seem logical that if there was a product or technology available that allows higher levels of storage availability to be achieved at the production site (or sites) it would be remiss not to consider them first before looking at the complexities of IT disaster recovery capabilities?
What if your critical data could exist in two places at once? geographically separated (whether by racks, rooms, building or states), managed as one and provide transparent IO fail-over and fail-back for all of your application servers?
Surely that would be worth considering?
DataCore Software’s core solution architecture is built around this capabilities, it is called “High Availability, synchronous mirroring” and provides the following benefits;
- physical separation of storage controllers
- physical separation of back-end disk systems
- Automated fail-over and fail-back
- is transparent to the application servers
- mirrored cache protection / consistency
