Date: Tue, 07 Jan 2003 12:31:29 -0700 From: "Justin T. Gibbs" <gibbs@scsiguy.com> To: R P <rpsbsd@cox.net> Cc: freebsd-scsi@freebsd.org Subject: Re: scsi parity errors Message-ID: <448790000.1041967888@aslan.btc.adaptec.com> In-Reply-To: <3.0.5.32.20030107115902.009423f0@pop.west.cox.net> References: <3.0.5.32.20030107063959.00916c40@pop.west.cox.net> <3.0.5.32.20030107063959.00916c40@pop.west.cox.net> <3.0.5.32.20030107115902.009423f0@pop.west.cox.net>
next in thread | previous in thread | raw e-mail | index | archive | help
>> How is the controller terminated? Are you using auto termination for >> both the primary (LVD only) and secondary (LVD/SE) bus segments? > > Yes. The current config is two drives on the LVD and only the CDRW on the > 50 pin SE. Nothing is on the 68 pin SE. Okay, then the manual termination configuration I mentioned before will not work. I assumed that all three connectors were in use. >> If >> you boot the system with verbose set, the aic7xxx driver will indicate >> how it has setup your controller termination. You need termination >> enabled on the primary segment and only the high byte termination enabled >> on the secondary segment. > > Verbose? How's that set? Hit a key during boot so you are dropped into the loader's "OK" prompt. Issue the command "boot -v". > Setting Low off and high on resulted in no CDRW detected by the > controller. Expected since, as you mentioned above, you have now removed the termination from the 68pin SE connector. You need to enable high and low byte termination on the secondary port if you are not providing it via an external/cable/ device terminator. > Even with both the SE cables disconnected, Again, high and low need to be on in this configuration. > Are you saying that in this case, the drives on the ends of the 50 pin and > 68 pin SE cables should not terminate the cables? All cables must be terminated at their ends. What I'm saying is that you may need to disable some of the termination on the controller depending on your cable configuration. The auto setting should work, but I will only know if the controller got this correct once I've seen the verbose boot messages along with a description of what cables were connected during that boot. >>> Windows NT, and 2000 using supplied drivers from Adaptec and no drive >>> parity or other drive errors were detected. >> >> The non-FreeBSD drivers perform domain validation and will silently >> perform speed fallback to avoid parity errors. > > One cable with terminator, one unterminated drive attached, controller > termination set to automatic, many parity errors. What's left? Bad cable > or Card? Probably a bad cable, bent pin, etc. -- Justin To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?448790000.1041967888>