Do most SAN storage vendors consider application support or operating system support when rating its HA solutions as "hot swappable"? (e.g. an OS can typically recover from the millisecond reset which sometimes occurs when an FC or SCSI disk is "hot swapped." However applications like Microsoft Exchange 5.5 will often see this reset as a timeout that will place the Exchange database and the Exchange logs out of sync and cause Exchange to drop off-line. Sometimes even marking the database as corrupt, thus forcing a restore of the database).
To guard against this possible database corruption, some SAN vendors will "silently" recommend that an application be shutdown or suspended while a disk is "hot swapped." Can these SAN solutions really be considered "highly available" if their components are really NOT "hot swappable" with respect to application support?
The short answer to your first question is no. Generally, vendors are only concerned about their own links in the chain when they discuss availability. They need to make themselves and their statistics look as good as possible when compared to their competition, so in general, they do not consider those other factors. Including external factors can only hurt the apparent goodness in their numbers and since those external factors are not their responsibility, there is no reason for them to be included. You wouldn't expect, for instance, a system vendor like Hewlett-Packard to count California's old rolling blackouts in their availability statistics or guarantees.
The answer to your somewhat philosophical second question is "it depends." My working definition of high availability is a business-based one. If your business' uptime requirements can be met by the components you have, then you have high availability. If they cannot, then you don't. No nines, no specific technologies, just a general business statement. So, for some businesses, the model that these vendors offer is highly available and for others, it is not.
Unfortunately, there is no formal definition of high availability so anyone can claim anything they want. Caveat emptor.
Hope this helps.
Evan L. Marcus
Editor's note: Do you agree with this expert's response? If you have more to share, post it in one of our .bphAaR2qhqA^0@/searchstorage>discussion forums.
Dig Deeper on SAN technology and arrays
Related Q&A from Evan Marcus
Storage expert Evan Marcus compares software and hardware RAID and outlines the benefits and drawbacks of each. Continue Reading
This expert answer explains the purpose of creating LUNs and details reasons for creating multiple LUNs. Continue Reading
This advice details the hardware and software requirements for setting up two data servers in fail-safe cluster mode for high availability. Continue Reading