Date: Sun, 12 Dec 1999 23:13:34 +0100 From: "Oliver Blasnik" <ob@omnilink.net> To: <freebsd-scsi@FreeBSD.ORG>, "Justin T. Gibbs" <gibbs@narnia.plutotech.com> Subject: Re: Again: CRD-Raid-Controller and FreeBSD 3.x Message-ID: <006201bf44ee$21a01df0$c20effd4@jav.net> References: <199912121537.IAA27060@narnia.plutotech.com>
next in thread | previous in thread | raw e-mail | index | archive | help
I asked after several problems: > > As a work-around the only thing was to disable tagged-queuing on all systems > > (added that devices to cam_xpt.c and build a new kernel). None of these > > messages occured anymore. Justin wrote: > The timeout should be 60 seconds. That is a veritable eternity for > a disk, especially since we do issue ordered tagged commads from > time to time to ensure that no tag starvation occurs on the device. Ok, 60 seconds should be enough :) > One thing you could try doing is limiting the number of tags we > attempt to use. Take a look at the quirk entries for the Quantum > Atlas drives in cam_xpt.c for details. I just checked out the datasheet-pdf of that 5440 at cmd.com. They write something about 32 commands-cache per host. Possibly they do mean tagq. Btw, I can en/disable tagged queuing at controller-level, but not set the size of the queue. That brings me to the question -> how much tagq's are default with that freebsd-drivers? If it's more than 32, a nice PR should be transmitted with including that CMD-Controllers to cam_xpt.c's list. > Did you try contacting CRD's technical support? Not yet, it's Sunday and I am @home... Cu, Oliver 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?006201bf44ee$21a01df0$c20effd4>