Date: Fri, 31 Jan 1997 05:54:03 -0500 (EST) From: Peter Dufault <dufault@hda.com> To: Shimon@i-Connect.Net (Simon Shapiro) Cc: julian@whistle.com, freebsd-scsi@freebsd.org Subject: Re: NewComer Questions... Message-ID: <199701311054.FAA12508@hda.hda.com> In-Reply-To: <XFMail.970130154549.Shimon@i-Connect.Net> from Simon Shapiro at "Jan 30, 97 12:44:23 pm"
next in thread | previous in thread | raw e-mail | index | archive | help
> > > 1. Multi-initiator support
> > I assume you mean in SCSI?
> > we ahve some basic support for that but it requires a SCSI host
> > adapter that supports it.. it hasn't been exercised in years.
> > (I wrote it iwith Peter Dufault but it's a rarely used feature.
> > Or are you alking about several machines sharing a single bus?)
>
> Yes. Multiple machines sharing the same SCSI bus. We have the HBA
> to do that and are porting the basic driver now. I am just
> wondering about what is there already and what is not.
Julian meant target mode and not multiple initiator, i.e., using
processor send/receive commands to talk between two processors over
the SCSI bus. That only works on the AHA1542B due to firmware
issues, though on that it worked fine a few years ago. I haven't
used it much since then.
I use a shared SCSI bus between two systems for debugging (again,
with 1542s; build using a fast system that stays up and reboot the
slow system) as you plan on doing and that works fine as long as
you stay on your toes and don't do anything stupid with multiply
mounted file systems.
Are you using the SCSI disk block range reservation in your DLM? I
thought about doing that but decided it was too likely to uncover
bugs in drive firmware and never even tried it.
It would certainly improve safety if it worked.
>
> > > 2. DLM
> > Daringly Lowfat Milk?
>
> You almost got it right... :-) But to be more precise, it stands
> for Distributed Lock Manager. A creature that is used in concert
> with multi-initiator SCSI busses and is responsible for providing
> the coordination necessary for such a mayhem. this arrangement is
> useful in two places: Large, complex databases, where more than
> one host (CPU, system) wants access to the same physical database
> and in HRA (High Reiliability and Availability) systems where one
> system failure still leaves a path to the database through another.
I assume you do this over ethernet and not processor send/receive?
(...)
>
> > > 2. Naming conventions for /dev entries for such beasts.
> > /dev/{r}sd[0-9][0-9]
> > there is a limit at the moment to about 32 drives per machine
> > (I think) but it's a rather artificial limit
> > and could be removed relativly easily.
>
> Good. How is this limit imposed? I need to know.
The field for units in the minor number is only 5 bits. The
dkunit stuff has a lower limit I think, but that affects only
disk statistics.
(...)
> When the initializing routines are running, are interrupts enabled?
> Is the VM totally functional? Can one sleep in the initializing?
No interrupts, no sleeping. The kernel memory is there and you can
malloc etc, but I'm not sure what you mean about VM totally functional.
--
Peter Dufault (dufault@hda.com) Realtime Machine Control and Simulation
HD Associates, Inc. Voice: 508 433 6936
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199701311054.FAA12508>