A6795A maps some LUNs, not others

[Phil Miesle_1]

Hello,

We've an A6795A which works great with our HP storage. This card resides in an rp7400 running HP-UX 11.11, and the driver is up-to-date. (I've not been able to find any evidence of a firmware update for this card?)

Now, I'm trying to attach another A6795A card to some Sun storage (a StorEdge 6960)...don't ask why :-) Sun tells me that their storage is compatible with this FC card on HP-UX 11.0; I don't see why it wouldn't work with 11.11...

The HP-UX system is able to see the LUNs...I know this because I can see them in "ioscan -fnk"...they appear as "disk" like so:

---
disk 17 1/12/0/0.16.136.255.0.0.1 disk CLAIMED DEVICE SUN SESS01
---
So it sees the LUN, except that there isn't a /dev device assigned...small problem, that.

*AND* (probably most telling of all) if I look at "dmesg" output (messages also seen at boot):

-----
DIAGNOSTIC SYSTEM WARNING:
The diagnostic logging facility has started receiving excessive errors from the I/O subsystem. I/O error entries will be lost until the cause of the excessive I/O logging is corrected. If the diaglogd daemon is not active, use the Daemon Startup command in stm to start it. If the diaglogd daemon is active, use the logtool utility in stm to determine which I/O subsystem is logging excessive errors.
----

I get this message from the card attached to the Sun storage but not the card attached to the HP storage. Something is amiss, I'm just not sure what.

Any ideas?

Thanks,
:-Phil

[Eugeny Brychkov]

Phil,
looks like your SAN storage needs to be configured. Current mode it uses is paripheral addressing mode, although it should use volume set addressing mode.
I do not know this storage, but by analogy with VA, HSG etc this array can have a table which assigns right array behavior for OS installed on particular host. For example, if you use w2k server array should represent LUN in w2k-like manner, if hpux, then in hpux-like manner etc. Do not ask what is the difference - its in the deeps of OS device handling
Eugeny

[Phil Miesle_1]

Eugeny,

Is this likely something that I can change at the OS-level for just this one card (i.e. not the one connected to the HP storage), or is it something that I *must* change on the SAN?

And if it is at the SAN level, why should this storage work okay under HP-UX 11.0 and not 11.11? Surely the device assignment logic hasn't change between the two OS releases?

Thanks,
:-Phil

[Eugeny Brychkov]

Are you sure this device will work under 11.0 in THIS configuration? Did you try? As I already said there's something wrong with array or SAN this array connected to. There can be nothing wrong with HBA, because it's autosensing.
What's topology in this SAN? Other devices? Which switch?
First of all you have to ensure that disk array is configured properly to work with hpux
Eugeny

[Eugeny Brychkov · Accepted solution]

Ok. Let's forget about disk arrays and SAN. You wrote that ioscan sees something but there're no device files. Try to install them using 'insf -e' command, and then issuing 'ioscan -fn' command to see if there're corresponding device files created.
The second is that you receive EMS notifications. Please use support tools manager ('mstm' for terminals, 'xstm' for graphics), go to Tools, then Utilities, then run 'logtool', select latest raw file, format it and view details. There you'll be able to see what's exactly these notifications mean and what's happening to the system
Eugeny

[Phil Miesle_1]

'insf -e' picked up the LUNs, so that did the trick there (thanks very much!)

I'm going to have to figure out a bit more with the logtool, but this is really a "me must learn more HP-UX stuff" than a SAN issue.

Upshot is that the "discovery" problem is solved...this should stay permanently in the configuration?

[Eugeny Brychkov]

Yes. Device files are created only once and live over reboots. If you connect device when system is up (like SAN disk array) you should make host see hardware (issuing 'ioscan -fn' hardware scanning command) and then create device files 'insf -e' for newly seen devices. But if you, after attaching device, reboot host, all these operations are done automatically during OS bootup.
After reviewing logtool logs please get back here with results/log contents
Eugeny

[Phil Miesle_1]

These error messages seem to be "typical" behaviour at reboot; the SAN gets similar messages but they go away once the host gets logged in...

Don't know why this is the case. I'm going to be working with an HP engineer to figure this out but this will be into next week before our local engineer can give a hand.

Thanks for all your help, I'll update this message when I learn something more!