Date: Tue, 18 May 2004 20:58:47 +0200 From: Jens Schweikhardt <schweikh@schweikhardt.net> To: "Justin T. Gibbs" <gibbs@scsiguy.com> Cc: freebsd-scsi@freebsd.org Subject: Re: 2120S Stripe - abysmal performance Message-ID: <20040518185847.GA3514@schweikhardt.net> In-Reply-To: <2991245408.1084458741@aslan.scsiguy.com> References: <FE045D4D9F7AED4CBFF1B3B813C85337051D8C0D@mail.sandvine.com> <40A26356.409@freebsd.org> <20040512183231.GB3137@schweikhardt.net> <20040512190429.GA1315@schweikhardt.net> <40A27607.9040801@freebsd.org> <20040512191949.GA1016@schweikhardt.net> <2991245408.1084458741@aslan.scsiguy.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Thu, May 13, 2004 at 08:32:21AM -0600, Justin T. Gibbs wrote: # > On Wed, May 12, 2004 at 01:07:51PM -0600, Scott Long wrote: # > ... # ># Something on the SCSI bus is very unhappy. What drives are they? What # ># firmware rev? # > # > 2 Fujitsu MAP3367NP with fw 0105. They've been working for a year # > on my 29160 without a hitch. # # At U320 speeds, I believe you need 0107 or better to have these # drives work. You should contact Fujitsu to ensure you have the # latest firmware. Okay, I contacted Fujitsu and now I have 0108 on both MAPs (quick answer from their German web support mail form, thanks Fujitsu!). Indeed my SCSI bus appears happy now. No more resets or other errors. I configured all IDs for U320 again. However, configuring a RAID 0 with both MAPs on the 2120S as the only drives still gives a sustained read rate of 65MB/s (constant over the whole stripe), while a single drive on the 29160 peaks at just over 70MB/s and drops to 40 at the inner cyls. I played around with the read cache, which has NO effect (unlike Scott explained elsethread). I also tried the TUNING tips from aac(4), e.g. hw.aac.iosize_max=96*1024 in /boot/loader.conf, also no noticeable change in performance. Should the 2120S be able to deliver more? With a vinum stripe I get 100MB/s. This is on a system with only 33MHz/32bit PCI (Asus A7N8X), and both of 29160 and 2120S. Somehow I get a feeling I'm hitting some limit that's just half the theoretical PCI bandwith of 133MB/s. Should I remove the 29160 (with just a slow CDROM) and try again? Is an update in order for the controller? AAC0> controller details Executing: controller details Controller Information ---------------------- Remote Computer: S Device Name: S Controller Type: No Info Access Mode: READ-WRITE Controller Serial Number: Last Six Digits = BA12C0 Number of Buses: 1 Devices per Bus: 15 Controller CPU: i960 R series Controller CPU Speed: 100 Mhz Controller Memory: 64 Mbytes Battery State: Not Present Component Revisions ------------------- CLI: 1.0-0 (Build #5263) API: 1.0-0 (Build #5263) Miniport Driver: 4.0-0 (Build #6011) Controller Software: 4.0-0 (Build #6011) Controller BIOS: 4.0-0 (Build #6011) Controller Firmware: (Build #6011) Controller Hardware: 2.64 Regards, Jens -- Jens Schweikhardt http://www.schweikhardt.net/ SIGSIG -- signature too long (core dumped)
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040518185847.GA3514>