- 
Mini Catasrophe averted
Category Win2k3
Today our SAN experienced a bit of a "burp" .... not sure what else to call it really. One of the drive shelves (that happens to be an entire LUN unto itself) decided it was going to unbind itself from its assigned storage processor, causing a whole lot of havoc (just happens to be one of two RAID 5 LUNs used by our main file server in a software-mirrored set for a bit of extra redundancy) ... Anyway, when it burped, I was able to bring the one of the two LUNs back online immediately, but the data was months old. (Note that prior to today, the software RAID mirrored set was reporting as healthy, synchronized and up to date).
After shutting down the SAN and rebooting everything, Good ol' WIn2k3 reported the it found both LUNs and it was healthy and all was right with the world. I didn't believe it, not after seeing months old data. So I broke the mirror, assigned drive letters to each LUN and took a look. Well, needless to say, it wasn't. In fact it looks like it had stopped synchronizing late 2007, but failed to mention this. Not only failed to mention it, it was reporting itself as HEALTHY. One LUN was up to date, the other, seriously behind. MONTHS behind. It's another WTF? moment brought to you by MS....How hard is it for an OS or a File system to realize its not synching properly. One more ding in the faith I'm "supposed" to have in the Windows OS.
Today our SAN experienced a bit of a "burp" .... not sure what else to call it really. One of the drive shelves (that happens to be an entire LUN unto itself) decided it was going to unbind itself from its assigned storage processor, causing a whole lot of havoc (just happens to be one of two RAID 5 LUNs used by our main file server in a software-mirrored set for a bit of extra redundancy) ... Anyway, when it burped, I was able to bring the one of the two LUNs back online immediately, but the data was months old. (Note that prior to today, the software RAID mirrored set was reporting as healthy, synchronized and up to date).
After shutting down the SAN and rebooting everything, Good ol' WIn2k3 reported the it found both LUNs and it was healthy and all was right with the world. I didn't believe it, not after seeing months old data. So I broke the mirror, assigned drive letters to each LUN and took a look. Well, needless to say, it wasn't. In fact it looks like it had stopped synchronizing late 2007, but failed to mention this. Not only failed to mention it, it was reporting itself as HEALTHY. One LUN was up to date, the other, seriously behind. MONTHS behind. It's another WTF? moment brought to you by MS....How hard is it for an OS or a File system to realize its not synching properly. One more ding in the faith I'm "supposed" to have in the Windows OS.
| Permalink | 
