Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!evax!utacfd!merch!cpe!adaptex!neese
From: neese@adaptex.UUCP
Newsgroups: comp.unix.sysv386
Subject: Re: <10127@nstar.uucp>
Message-ID: <77900008@adaptex>
Date: 10 Sep 90 16:34:00 GMT
References: <10127@nstar.uucp>
Lines: 30
Nf-ID: #R:nstar.uucp:10127:adaptex:77900008:000:1515
Nf-From: adaptex.UUCP!neese    Sep 10 10:34:00 1990


>>>If you use SCSI drives, at least some of them power-sequence themselves;
>>>that's a consideration since it's the start-up power requirement that's the
>
>>I tried enabling the delay on power up on my SCSI Wren drives - but
>>the Adaptec 1542B & ISC never finds the drives - 
>
>Curious.  I missed the first part of this thread, but it looks suspiciously
>like the behavior of my 1542B/ Seagate 296N combination under Xenix 2.3.2.
>On the initial power-up sequencing, the Adaptec will poll for devices
>before the Seagate has finished it's recalibration. After about ten seconds
>of silence, the ST296N will restart what sounds like it's power-on self
>test, and the 1542B will then report it as Drive C and continue.
>
>Come to think of it, this is not OS specific, because it does this when I
>run it under DOS as well.

What the behavior you are describing is quite simple.  At power on a SCSI
device goes through it's self-test.  Some SCSI devices will hold the SCSI
bus busy during this period of time.  These devices will cause a SCSI bus
reset to be issued by the adapter after the SCSI bus goes free so you will
see the device go through it's self test again.
If a OS issues a SCSI bus reset and does not wait for the bus to clear up
before going further, there is a good chance that they will not recognize the
device when they get to the first issuance of a SCSI command.

			Roy Neese
			Adaptec Senior SCSI Applications Engineer
			UUCP @  uunet!swbatl!
				  {nominil,merch,cpe,mlite}!adaptex!neese