[8358] in linux-scsi channel archive
Re: SD_EXTRA_DEVS (was Problem)
daemon@ATHENA.MIT.EDU (Kurt Garloff)
Fri Mar 10 14:50:20 2000
Date: Fri, 10 Mar 2000 17:55:59 +0100
From: Kurt Garloff <garloff@suse.de>
To: Alan Cox <alan@lxorguk.ukuu.org.uk>
Message-ID: <20000310175559.B2340@mobil.tue.nl>
Mime-Version: 1.0
Content-Type: multipart/signed; micalg=pgp-md5;
protocol="application/pgp-signature"; boundary="RIYY1s2vRbPFwWeW"
In-Reply-To: <E12TQCo-0003Bq-00@the-village.bc.nu>; from alan@lxorguk.ukuu.org.uk on Fri, Mar 10, 2000 at 02:15:54PM +0000
--RIYY1s2vRbPFwWeW
Content-Type: multipart/mixed; boundary="5I6of5zJg18YgZEa"
--5I6of5zJg18YgZEa
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: quoted-printable
On Fri, Mar 10, 2000 at 02:15:54PM +0000, Alan Cox wrote:
> Ok I picked roughly similar numbers. Can you tell me where you fished the=
21
> out from and why (just curious)
I think I made it up somehow like this:
Going out from the current numbers, I was thinking about how many external
devices of that kind people might want to connect and ended up with smth
like 20, 3, 8 for sd, st, sr.
20 sd devs are 4 arrays =E0 5 disks
8 sr devs is a changer which uses 7 LUNs plus a CD-R
3 st devs: most changers just use one LUN for the tape access and as st
devs take up by far the most mem ...
=20
This would result in 31 sg devs, but I like a number like 32 much better ...
so I increased one of the numbers by one. I think the tapes took the largest
amount of memory, so I added a sd dev. People with huge SCSI installations
mostly do it in order to connect a lot of disks (RAID arrays normallyy), not
because they have 8 tape devices, IME.
I don't have my notebook (one out of paper ...), where I put the notes for
the calculation of the needed mem that time, here at Univ, but I can look it
up, when I'm home again ... (but where I don't currently have mail :-(
> I'll probably change the 2.2.15 ones to match the SuSE ones. Mine are sim=
ilar
> values and we might as well just have one default size of extra devices
If you came to similar values, that indicates that they are in the region
for a reasonable compromise between saving resources (mem) and allowing for
easy extendability of the SCSI subsystem.
Of course, I appreciate, if you use the same numbers.=20
I attach the trivial patch ...
Regards,
--=20
Kurt Garloff <kurt@garloff.de> [Eindhoven, NL]
Physics: Plasma simulations <k.garloff@phys.tue.nl> [TU Eindhoven, NL]=20
Linux: SCSI, Security <garloff@suse.de> [SuSE Nuernberg, FRG]
(See mail header or public key servers for PGP2 and GPG public keys.)
--5I6of5zJg18YgZEa
Content-Type: text/plain; charset=us-ascii
Content-Disposition: attachment; filename="scsi-moreextradevs.diff"
--- linux-2.2.13p16.work/drivers/scsi/hosts.h.orig Tue Oct 12 19:48:18 1999
+++ linux-2.2.13p16.work/drivers/scsi/hosts.h Wed Oct 13 12:32:11 1999
@@ -484,10 +484,10 @@
#ifdef CONFIG_SCSI_PLUTO_MODULE
#define SD_EXTRA_DEVS 40
#else
-#define SD_EXTRA_DEVS 4
+#define SD_EXTRA_DEVS 21
#endif
-#define ST_EXTRA_DEVS 2
-#define SR_EXTRA_DEVS 2
+#define ST_EXTRA_DEVS 3
+#define SR_EXTRA_DEVS 8
#define SG_EXTRA_DEVS (SD_EXTRA_DEVS + SR_EXTRA_DEVS + ST_EXTRA_DEVS)
#endif
--5I6of5zJg18YgZEa--
--RIYY1s2vRbPFwWeW
Content-Type: application/pgp-signature
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.1c- (GNU/Linux)
Comment: For info see http://www.gnupg.org
iD8DBQE4ySkexmLh6hyYd04RAZBlAJ0ehkoMFXU7LHQUb/GTF8mWXc/0owCguSqB
Zfvlqru+NUtBg5zkhCdgwuU=
=gmCA
-----END PGP SIGNATURE-----
--RIYY1s2vRbPFwWeW--
-
To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@vger.rutgers.edu
