Date: Mon, 22 Feb 1999 09:22:32 -0500 (EST) From: Andrew Gallatin <gallatin@cs.duke.edu> To: Ruslan Ermilov <ru@ucb.crimea.ua> Cc: "Kenneth D. Merry" <ken@plutotech.com>, stable@FreeBSD.ORG, allenc@verinet.com Subject: Re: Tagged queueing Message-ID: <14033.25751.508511.815083@grasshopper.cs.duke.edu> In-Reply-To: <19990222140147.A2015@ucb.crimea.ua> References: <19990219233927.A4953@ucb.crimea.ua> <199902201930.MAA13801@panzer.plutotech.com> <19990222140147.A2015@ucb.crimea.ua>
next in thread | previous in thread | raw e-mail | index | archive | help
Ruslan Ermilov writes: > On Sat, Feb 20, 1999 at 12:30:43PM -0700, Kenneth D. Merry wrote: > > > > The AHC_TAGENABLE option is no longer necessary. Tagged queueing is > > enabled by default. > > > > The reason your drive does not have tagged queueing enabled is because it is > > a Western Digital Enterprise drive: > > da0 at ahc0 bus 0 target 0 lun 0 > > da0: <WDIGTL ENTERPRISE 1.80> Fixed Direct Access SCSI-2 device > > da0: 40.000MB/s transfers (20.000MHz, offset 8, 16bit) > > da0: 4157MB (8515173 512 byte sectors: 64H 32S/T 4157C) > > > > Andrew Gallatin reported that his western digital enterprise drives only > > get 1.5MB/sec throughput with tagged queueing turned on, and 8MB/sec > > throughput with tagged queueing turned off. Therefore, we have a quirk > > entry in the transport layer that disables tagged queueing for those > > drives. > Reported? What's the number of his problem report? I talked to him via private email shortly after CAM went into -current. I have two WDE's: da0 at ahc0 bus 0 target 0 lun 0 da0: <WDIGTL WDE4360-1807A3 1.80> Fixed Direct Access SCSI-2 device da0: 40.0MB/s transfers (20.0MHz, offset 8, 16bit) da0: 4095MB (8388314 512 byte sectors: 255H 63S/T 522C) da1 at ahc0 bus 0 target 1 lun 0 da1: <WDIGTL ENTERPRISE 1.91> Fixed Direct Access SCSI-2 device da1: 40.0MB/s transfers (20.0MHz, offset 8, 16bit) da1: 4157MB (8515173 512 byte sectors: 255H 63S/T 530C) Both show piss-poor performance with TQ enabled. Note that they are 'different' than yours at least in that my drive with 4157MB has revision 1.91 of the firmware. Perhaps that has something do do with it... Anyway, with TQ disabled, WCE on both drives I see decent performance: <9:12am>grasshopper/gallatin:tmp>df /var/tmp Filesystem 512-blocks Used Avail Capacity Mounted on /dev/da0s1e 575486 453702 75746 86% /var <9:12am>grasshopper/gallatin:tmp>dd if=/dev/zero of=/var/tmp/test bs=64k count=1000 1000+0 records in 1000+0 records out 65536000 bytes transferred in 8.440735 secs (7764253 bytes/sec) <9:13am>grasshopper/gallatin:tmp>df /freebsd/tmp/ Filesystem 512-blocks Used Avail Capacity Mounted on /dev/da1s1c 8252330 4384130 3208014 58% /freebsd <9:13am>grasshopper/gallatin:tmp>dd if=/dev/zero of=/freebsd/tmp/test bs=64k count=1000 1000+0 records in 1000+0 records out 65536000 bytes transferred in 7.876551 secs (8320393 bytes/sec) <9:15am>grasshopper/gallatin:tmp>uname -a FreeBSD grasshopper.cs.duke.edu 4.0-CURRENT FreeBSD 4.0-CURRENT #0: Sun Jan 24 16:17:20 EST 1999 gallatin@grasshopper.cs.duke.edu:/freebsd/src/sys/compile/GRASSHOPPER i386 I'm sorry, but I've got a bit too much state laying around to be able reboot to turn on tagged queuing to verify my drives performance with TQ enabled still stinks... Drew ------------------------------------------------------------------------------ Andrew Gallatin, Sr Systems Programmer http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?14033.25751.508511.815083>