linux-kernel - newer MPT speed issue on Linux

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Message-ID: <44BD1AB6.10009@atipa.com>
Date:	Tue, 18 Jul 2006 12:30:30 -0500
From:	Roger Heflin <rheflin@...pa.com>
To:	Linux-Kernel <linux-kernel@...r.kernel.org>,
	mpt_linux_developer@...l.com
Subject: newer MPT speed issue on Linux

Hello,

I am have tested with Sles9.1sp1, Sp3, and kernel.org 2.6.17.6 and
here are some results:

SLES9.1sp1 (kernel only)	100-120MB/second writes  75-80MB/second reads
SLES9.0sp3			20-30MB/second writes
2.6.17.6			20-30MB/second writes   140-150MB/second reads

This is with  the LSI Logic / Symbios Logic FC929X Fibre Channel Adapter 
(rev 81)

All 3 are with a SLES9sp3 distribution and are on the same machine with 
only the kernel
being changed.   The test is a simple "dd if=/dev/zero of=/dev/dk001/two 
bs=65536 &"
with vmstat being used to determine io speed.   Bonnie++ is getting 
similar results
to what dd is generating when writing to a filesystem.

The fiber channel card is hooked to a external raid5 chassis, and it 
being accessed
though LVM.

Are there some parameters that I am missing that could speed the writes up?

The reads on the newer driver are a nice improvement, but the writes on
the newer driver are horrible in my testing.

                               Roger
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/