Date: Wed, 04 Mar 1998 16:09:06 -0800 (PST) From: Simon Shapiro <shimon@simon-shapiro.org> To: Wilko Bulte <wilko@yedi.iaf.nl> Cc: grog@lemis.com, hackers@FreeBSD.ORG, blkirk@float.eli.net, jdn@acp.qiv.com, tlambert@primenet.com, sbabkin@dcn.att.com Subject: Re: SCSI Bus redundancy... Message-ID: <XFMail.980304160906.shimon@simon-shapiro.org> In-Reply-To: <199803042207.XAA04235@yedi.iaf.nl>
next in thread | previous in thread | raw e-mail | index | archive | help
On 04-Mar-98 Wilko Bulte wrote: ... > For good performance our rule of thumb is 4-5 disks/bus. Matches your's > quite nicely. I hope others are reading this. I hear conflicting opinions, quite argumentatively, all the time. I hope I do not offend anyone, but these numbers were true about 20 years ago, and somehow, they stay constant, or get worse with time. This is partially why when the SASI spec turned into SCSI, the number of address lines went down from 8 to 3. The wieres are still there :-) > 3 buses minimum is more based on the rule that you don't want to have > more than > one disk out of each set on a channel. As Murphy has it it is always the > bus > with >1 disk that somehow gets killed. A RAID5 of 3 disks is pretty > minimal > (OK, it might be 36 Gb netto cap now) Yup. If you stick with RAID-1, you can get away with 2 busses. Now, connecting a RAID box like this to your computer with exactly ONE SCSI cable is .... ... > OK, I understand you intend to compete with NT/Wolfpack (OK, MS Cluster Server they call it now I think). Not really. I just need this functionality, belive many others do too, and want to assure NT does not get them because Unix does not have it, or has it on proprietary hardware for more money per system than I make in a year. > What do you call it? 'ChuckPack' ? ;-) It is traditional with my people not to name a child until he/she is born and proven viable and healthy :-) ... > You are aiming for second-failover times? How do you distinguish then > between a somewhat slow machine and one that is really dead? Watchdog timers, mainly. SCSI-SCSI communications, secondary, RS-232 modem control lines tricery. The WDTs are cross wired. Kernel facilities to attach callbacks to events are already there and running. This entire conversation belongs in databases. No? ---------- Sincerely Yours, Simon Shapiro Shimon@Simon-Shapiro.ORG Voice: 503.799.2313 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?XFMail.980304160906.shimon>