On Jun 7, 2010, at 6:58 AM, Richard Rhodes wrote:

> I'm not sure that a dismantle will solve your problem, but your problem 
> sounds similar to a couple situations we've had where it did resolve the 
> problem.
> 
> Your SAN consultant is correct - it shouldn't be necessary!

I don't know about anyone else, but, I don't believe in the word "should" when 
it comes to IT, and I don't trust my SAN clients that I can't control. Even the 
ones I control, I double-check at the first sign of trouble.

SAN clients' fibre drivers do dumb things, especially in the realm of SAN 
discovery or the consistency of how they present devices to applications like 
TSM. The shop I was in a year-and-a-half ago had a months-old crit-sit that was 
old when they hired me and still not quite dead when I was laid off mere months 
later. That one was at least partly caused by a Windows DB server that couldn't 
disable SAN discovery to save its life, and that kept moving known devices to 
new addresses. In that case, we eventually zoned that particular SAN client so 
it was isolated to just a few tapes so its impact was limited. Until we figured 
that out, there were lots of power-on-resets of LTO drives in a real 3584 that 
were tainted by this SAN client as it tried to find the drives it was being 
told to use. 

With virtual tape libraries, the symptoms are a bit more manageable, but SAN 
clients still get clumsy and stomp all over everything. 

I'm not saying it's the root cause here, but if it is, you might here me 
mumble, "Been there, done that." 

Good luck,
Nick