Date: Wed, 11 Nov 1998 15:16:28 +1030 From: Greg Lehey <grog@lemis.com> To: Julian Elischer <julian@whistle.com>, Peter Jeremy <peter.jeremy@auss2.alcatel.com.au> Cc: hackers@FreeBSD.ORG Subject: Re: SCSI tagged queueing and softupdates Message-ID: <19981111151628.J20374@freebie.lemis.com> In-Reply-To: <Pine.BSF.3.95.981110201400.15329G-100000@current1.whistle.com>; from Julian Elischer on Tue, Nov 10, 1998 at 08:16:08PM -0800 References: <98Nov11.084659est.40344@border.alcanet.com.au> <Pine.BSF.3.95.981110201400.15329G-100000@current1.whistle.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tuesday, 10 November 1998 at 20:16:08 -0800, Julian Elischer wrote: > On Wed, 11 Nov 1998, Peter Jeremy wrote: > >> AFAIK, SCSI tagged queueing is the default in 3.0 (via cam or inside >> the ahc driver). (This was optionally enabled via AHC_TAGENABLE in 2.x >> and the 3.0 man page still describes this option, although the option >> is gone). >> >> According to the description, tagged queueing allows a SCSI device to >> re-order pending I/O requests in order to improve throughput. Soft >> updates, on the other hand, rely on the correct sequencing of physical >> writes to ensure that the disk metadata is consistent. >> >> Is this a real issue? If so, it would seem that tagged queueing must >> be disabled (at least for writes) when softupdates are enabled. It's >> not obvious to me that this is currently done (or even easy to do). >> >> Comments please. > > transactions can be reorderd AS LONG AS the drive does not report a > perticular transaction to be completed before it really is.. As long as you define "transaction" correctly. SCSI-level transactions must often be completed in order. I rely on this (and the associated B_ORDERED flag in the buffer header) to ensure consistency in Vinum. > Soft updates will not schedule any IO that depends on an uncompleted IO > operation. So therefore all operations that have been 'passed to the > driver' are not dependent on each other (by definition) Doesn't this slow the whole thing down? Or is that what's behind B_ORDERED? Greg -- See complete headers for address, home page and phone numbers finger grog@lemis.com for PGP public key 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?19981111151628.J20374>