Date: Tue, 27 Jul 1999 08:14:11 -0700 (PDT) From: Matthew Jacob <mjacob@feral.com> To: "Justin T. Gibbs" <gibbs@plutotech.com> Cc: "Kenneth D. Merry" <ken@plutotech.com>, scsi@freebsd.org Subject: Re: delay in debugging problems with Qlogic cards caused by... Message-ID: <Pine.BSF.4.05.9907270811100.31194-100000@semuta.feral.com> In-Reply-To: <199907271508.JAA22101@caspian.plutotech.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> >It strikes me that an upper bound of busy time would be a good thing to > >have- it took me several days of hangs in the middle of the night until I > >finally caught this. You can say that this is "flaky" h/w (and you'd be > >right), but seems like a good case for defensive programming. Whew. I'm > >sore fatigued... > > What is a good upper bound on the length of time a device should be allowed > to report busy? > > I hate these kinds of questions because you never seem to be able to > satisfy everyone. Perhaps 5 minutes is good. 5 minutes is probably too short, but I most emphatically agree that it's not an easy question to answer- it's the same level as "How long should a command take" for devices like tapes. It probably should be a per-device property and quirked to override it- mostly so that the da driver can finally do a BDR or *something* to try and get the thing back alive- or at least complain about it... > One thing I don't understand yet is why this hung up the whole machine. > It should have only stuffed up that one device, not all devices in the > system. Have you looked at the code path to see where this other bug > may be? I'm sorry- I was a little inaccurate here- the machine would ping, but because it had the root/swap on it, nothing else could happen. -matt > 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?Pine.BSF.4.05.9907270811100.31194-100000>