Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!mips!pacbell.com!decwrl!pa.dec.com!granite.pa.dec.com!ajc From: ajc@thendara.pa.dec.com (AJ Casamento) Newsgroups: comp.unix.ultrix Subject: Re: daisy chaining SCSI devices on a ds3100 Message-ID: Date: 15 May 91 04:41:25 GMT References: Sender: news@pa.dec.com (News) Organization: Digital Equipment Corporation Lines: 78 In-Reply-To: robm@ataraxia.Berkeley.EDU's message of Fri, 10 May 1991 23:54:25 GMT Rob McNicholas writes: I have a DECstation 3100 (24M memory, ULTRIX V4.1 (Rev. 55), one storage expansion box) with the following SCSI configuration: SCSI id device order in chain ------- ------ -------------- 0 rz57 1 ? (internal to storage expansion) 1 rz55 3 2 rz57 1 ? (internal to storage expansion) 3 UNUSED UNUSED 4 TZK08 (exabyte 8mm) 2 5 TK50 4 I was wondering if the order the devices are connected together affects their performance. There are a few items to be aware of: * SCSI ID indicates priority, with the highest number receiving the highest priority. * Within the 6M limit of electrical cable equivalent, the devices should be impacted by their SCSI ID and not by the order from the system SCSI port. * SCSI Host ID - On the DECstation 2100/3100 systems this will be a default of 6 (the ID 7 was reserved for a Realtime CPU slot). On the DECsystem 5000 series (1XX or 2XX) it will default to ID 7). * In the first RISC/Ultrix SCSI driver implementations the following recommendations for ID assignments was made (I haven't seen it contradicted since) 7 Realtime CPU 6 CPU 5 TK50 4 RRD40 3 TZK08 2 RZ23 1 RZ55 0 RZ56 Obviously, there isn't a Realtime CPU so that address is useable. The idea is to give the slower SCSI devices (except for CPU) a higher SCSI ID number. In the case of the TK50 this is supposed to be a real help. The RRD40 is somewhat faster. The position of the 8mm is arguable. Although, I prefer to keep the tape media at a higher priority. Most of the disk drives should play well regardless of how you mix them. I will tell you honestly that I remember some flakey things happening when a disk drive was given a higher priority than the CPU, but I don't know if that's still true. I hope that this information helps. Thanx, AJ ********************************************************************** * AJ Casamento "The question is not whether or * * Digital's TRI/ADD Program not the opinions are mine; but * * 100 Hamilton Ave. UCO1-B rather, which of my personalities * * Palo Alto, CA 94301-1616 do they belong to?" * * 415.853.6744 * * ajc@decwrl.dec.com * **********************************************************************