SEO

I have a T3 on an FC loop (HUB) share by different hosts.
One of the hosts is a Blade 1000.
Other hosts are PCs with QLA2200F HBA.

While debugging this issue I have tried both with default QLA2200
firmware settings, and with a fixed HBA ID to each HBA. Here for
simplicity I will refer to fixed HBA IDs (0-125), although the problem
occurs either with a fixed ID or with autoconfiguration.

When I boot another host (PC + QLA2200F with BIOS enabled), the PC's
HBA (QLA2200) scans for devices on the loop and I see that stops right
at the address assigned to the Blade's HBA (100). At this point Blade
1000 logs FC loop access errors on the console and hangs there.
Sometimes it locks, sometimes it reboots, but it never recovers
resuming access to the loop (and to the internal FC drive that sits on
the same loop).
On the other end the HBA in the PC comes up with and empty list of
devices discovered.

Clearly something goes wrong when the QLA2200 in the PC discovers the
loop at boot (loop reset?). I have tried different combinations of
parameters (LIP on/off, target reset on/off, etc.) with no joy.

The opposite works most of the time (boot PC first, and then blade),
however sometimes blade hangs at boot, and requires two or three
attempts before a successful start-up, but the PC on the other end
(running Linux) complains about a devices access, but only
temporarily, resuming access after a few seconds. There is to note<br