home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: comp.unix.bsd
- Path: sparky!uunet!convex!darwin.sura.net!jvnc.net!yale.edu!nigel.msen.com!caen!hellgate.utah.edu!fcom.cc.utah.edu!gateway.univel.com!gateway.novell.com!terry
- From: terry@npd.Novell.COM (Terry Lambert)
- Subject: Re: SCSI tape drive not accessible
- Message-ID: <1992Jul28.171521.9397@gateway.novell.com>
- Sender: news@gateway.novell.com (NetNews)
- Nntp-Posting-Host: thisbe.eng.sandy.novell.com
- Organization: Novell NPD -- Sandy, UT
- References: <TMH.92Jul23041936@keks.first.gmd.de> <1992Jul23.080507.29145@news.uni-stuttgart.de> <5939@pdxgate.UUCP>
- Distribution: comp
- Date: Tue, 28 Jul 1992 17:15:21 GMT
- Lines: 40
-
- In article <5939@pdxgate.UUCP> brians@eecs.cs.pdx.edu (Brian Smith) writes:
- >In article <1992Jul23.080507.29145@news.uni-stuttgart.de> zrgl0376@helpdesk.rus.uni-stuttgart.de (Cowboy aka Bjoern Lemke) writes:
- >>In article <TMH.92Jul23041936@keks.first.gmd.de> tmh@keks.first.gmd.de (Thomas Hoberg) writes:
- >>>I managed to get the "Tiny BSD" floppy accross the modem and got
- >>>386BSD installed on my system (486/33, 16MB, ET4000/8514, AHA1542, 1GB
- >>>Wren VII + 88MB Syquest SCSI drives, Tandberg TDC3800 QIC 525 SCSI
- >>>Streamer). The SCSI disks are at targets 0+1, the tape drive is at
- >>>target 4. On boot up, only the first disk is recognized. Trying to
- >>
- >>I run into a similar problem. I used a Fujitsu 2624 Harddisk with the same
- >>Streamer (Tandberg). Setting the tape drive to target 4, resulted in
- >>nonrecognizing the streamer. (My harddisk is set to target 0). But after
- >>setting the tape device to target 1, the scsi controller got it and
- >>i was able to read and write from the device (/dev/ras1a) including some
- >>scsi error messages (sense 0 0 0 0 .. and so on). I am not a scsi guru
- >>and I don't know if this method is proper, but for the moment it works.
- >
- >This is in the faq recently posted. Apparently, the scsi driver in
- >386BSD only probes targets until it finds an empty spot. So, if you had
- >target 0, target 1 and target 4, then the driver would find target 0, then
- >target 1, then look for target 2 (finding it empty/non-responsive), and
- >conclude that no more devices were on the scsi bus. The workaround is to
- >make your scsi devices all contiguous and starting at 0.
-
- Uh, this is *wrong* in the FAQ... 8-(. I guess that's what I get for copying
- from other peoples papers. The problem is not that the devices are not
- recognized after one empty slot; it's that the device nodes are created
- incorrectly. Do an 'ls -l' and you will see obviously bogus minor device
- numbers. I have seen two instances correcting me; one was mailed direct,
- and one was posted here. Since no one has come forward to claim official
- keeper-of-the-FAQ status, I guess I will have to do at least one more posting
- of an updated version. Look for it some time this week.
-
-
- Terry Lambert
- terry_lambert@gateway.novell.com
- terry@icarus.weber.edu
- ---
- Disclaimer: Any opinions in this posting are my own and not those of
- my present or previous employers.
-