Most all SCSI disks claim to support queueing. Many of them have queueing
that works
wrong; if the tags are all treated alike (too common) things will work just
fine till you
get a cluster state transition. Then watch out. If they don’t accept all tag
values, things
also get screwed up and can hang the SCSI bus worst case. I’ve seen and
diagnosed
that case in a common brand of drives. There are numerous other things that
can go
wrong, but I have no interest in providing a rogues’ gallery of firmware
bugs here. Last
list I saw ran for several pages, couple of lines to a bug\\misfeature.
Yes, a VMS port to IA64 is in progress. There was prelim work on x86 some
years
ago. There is no guarantee what will happen, but the IA64 port will leave
the VMS
system much easier to move to other chips. That part of this happens by
editing a
compiler backend is unusual, but not a barrier once the requisite skill
level is reached.
re scsi errors:
Queue full or busy do not cause corruption, but they are not rare error
conditions either.
New SCSI drives, as well as the older ones, do occasionally have errors
though. A robust
I/O system needs to be able to respond predictably when these arise.
-----Original Message-----
From: Maxim S. Shatskih [mailto:xxxxx@storagecraft.com]
Sent: Monday, April 29, 2002 4:32 PM
To: NT Developers Interest List
Subject: [ntdev] Re: Philosophical Rant [was Re: Writing Drivers in Ja va]
model which most devices are unable to use in full generality. Thus unless
the SCSI devices really use tagged queueing and multi-initator code
*right*
IIRC all commodity SCSI disks support queuing.
the idea is not to use tagged queueing at all, lest rare error conditions
cause data corruption).
Will SCSISTAT_BUSY or SCSISTAT_QUEUE_FULL corrupt any data?
Of course, VMS has been a 64 bit system for some time now, on alpha (and
soon on IA64 and possibly Hammer after…)
They really want to port VMS on IA64 and Hammer? The latter is most amazing,
since Hammer is known to be direct extension of x86 to
64bits, not so with IA64 which is VLIW processor and has really nothing
common with x86.
Max
You are currently subscribed to ntdev as: xxxxx@FirstUSA.com
To unsubscribe send a blank email to %%email.unsub%%
**********************************************************************
This transmission may contain information that is privileged, confidential and/or exempt from disclosure under applicable law. If you are not the intended recipient, you are hereby notified that any disclosure, copying, distribution, or use of the information contained herein (including any reliance thereon) is STRICTLY PROHIBITED. If you received this transmission in error, please immediately contact the sender and destroy the material in its entirety, whether in electronic or hard copy format. Thank you.
**********************************************************************