Date: Fri, 22 Sep 2000 15:12:20 -0700 From: Mike Smith <msmith@freebsd.org> To: Bernd Walter <ticso@cicely5.cicely.de> Cc: Marius Bendiksen <mbendiks@eunet.no>, freeBSD-scsi@freebsd.org Subject: Re: disable write caching with softupdates? Message-ID: <200009222212.e8MMCKA00399@mass.osd.bsdi.com> In-Reply-To: Your message of "Sat, 23 Sep 2000 00:06:05 %2B0200." <20000923000605.A448@cicely5.cicely.de>
next in thread | previous in thread | raw e-mail | index | archive | help
> On Thu, Sep 21, 2000 at 10:10:06AM -0700, Mike Smith wrote: > > > > > OK, I played a bit with that, the only info I can see I get from the > > > > > higher levels is the BIO_ORDERED bit, so I tried to flush the cache > > > > > each time I get one of those, _bad_ idea, 10% performance loss... > > > > > > > That's the price of having a recoverable file system. See Seltzer, Ganger, > > > > > > Not necessarily. > > > > Er, you're being both contrary and plain wrong. It's a fundamental > > assumption of the softupdates implementation that it is possible to issue > > an ordered write and have it complete in an ordered fashion. > > Now I'm very confused ;( > > That's what Matthew Dillon wrote on -current in May: > > : Wait a sec... softupdates does not depend on write ordering. Softupdates > : issues all non-conflicting writes in parallel and doesn't care what order > : they are written to the disk. When those complete, softupdates will then > : followup with all the writes that depend on the original set. > > As far as I know Kirk McKusick had bad expiriences with disks not > honouring the ordered stuff. > > What is the truth? I stated it poorly. Softupdates does depend on the equivalent of "write barrier" semantics, as Matt describes. It doesn't explicitly issue "ordered" writes because it's running lots of separately dependant write chains in parallel, while a true "barrier" would apply across all writes. The semantic that softupdates desires is difficult to achieve. It can be best approximated using a disk controller with nonvolatile cache memory (as this reduces some of the windows to nearly zero). -- ... every activity meets with opposition, everyone who acts has his rivals and unfortunately opponents also. But not because people want to be opponents, rather because the tasks and relationships force people to take different points of view. [Dr. Fritz Todt] 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?200009222212.e8MMCKA00399>