Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 27 Mar 2014 10:42:37 +0100
From:      Borja Marcos <borjam@sarenet.es>
To:        Doug Ambrisko <ambrisko@cisco.com>
Cc:        "scottl@netflix.com" <scottl@netflix.com>, "Radford, Adam" <Adam.Radford@lsi.com>, "sean_bruno@yahoo.com" <sean_bruno@yahoo.com>, "Mankani, Krishnaraddi" <Krishnaraddi.Mankani@lsi.com>, "dwhite@ixsystems.com" <dwhite@ixsystems.com>, "Maloy, Joe" <Joe.Maloy@lsi.com>, "jpaetzel@freebsd.org" <jpaetzel@freebsd.org>, "freebsd-scsi@freebsd.org" <freebsd-scsi@freebsd.org>, "Kenneth D. Merry" <ken@kdm.org>, "McConnell, Stephen" <Stephen.McConnell@lsi.com>
Subject:   Re: LSI - MR-Fusion controller driver <mrsas> patch and man page
Message-ID:  <779F5D6D-BCA2-4376-84BC-6B08A8F357D4@sarenet.es>
In-Reply-To: <20140325174604.GA41527@cisco.com>
References:  <20140124185356.GA28724@ambrisko.com> <20140124190047.GA34975@ambrisko.com> <9c3fd2b15e9b4c2cb967519a3b7f98ad@BN1PR07MB247.namprd07.prod.outlook.com> <20140318143738.GA65955@cisco.com> <20140320235534.GA92797@cisco.com> <C698AC2A-06A7-4408-9790-20B69FAF31C6@sarenet.es> <20140321160954.GB99545@cisco.com> <5C32A3C7-B28B-4E69-9DF0-EE53181085F7@sarenet.es> <20140324174519.GA30345@cisco.com> <AE6FAC75-D2EA-457F-8654-2ECD57EDAA13@sarenet.es> <20140325174604.GA41527@cisco.com>

next in thread | previous in thread | raw e-mail | index | archive | help

On Mar 25, 2014, at 6:46 PM, Doug Ambrisko wrote:

> Quick question, what version of FreeBSD are you using?  I netbooted
> current and then did:
> 	newfs /dev/da4
> 	mount /dev/da4 /mnt
> 	cd /mnt ; bonnie++ -u 0
> da4 is a Toshiba 300G SAS drive.  This worked okay.

Sorry for the delay, I had to rearrange some stuff to repeat the tests.

For me, it corrupts data with -CURRENT, -STABLE (the one which has a =
backport of the mfi driver from -CURRENT) and 10-RELEASE.

I noticed (I didn't check it when I ran the first tests, as data =
corruption was enough of a problem) that the controller is negotiating =
just
150 MB/s. Not sure if the driver is lying, though.=20

# camcontrol devlist
<ATA Samsung SSD 840 AB0Q>         at scbus0 target 3 lun 0 (da0,pass0)
<ATA Samsung SSD 840 BB0Q>         at scbus0 target 4 lun 0 (da1,pass1)
<IBM-ESXS SAS EXP BP 61A6>         at scbus1 target 8 lun 0 (ses0,pass2)
<IBM-ESXS SAS EXP BP 61A6>         at scbus1 target 12 lun 0 =
(ses1,pass3)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 15 lun 0 (da2,pass4)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 16 lun 0 (da3,pass5)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 17 lun 0 (da4,pass6)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 32 lun 0 (da5,pass7)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 33 lun 0 (da6,pass8)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 34 lun 0 (da7,pass9)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 35 lun 0 =
(da8,pass10)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 36 lun 0 =
(da9,pass11)
<ATA Samsung SSD 840 BB0Q>         at scbus1 target 37 lun 0 =
(da10,pass12)
<SEAGATE ST9146803SS FS03>         at scbus1 target 39 lun 0 =
(da11,pass13)
<SEAGATE ST9146803SS FS03>         at scbus1 target 40 lun 0 =
(da12,pass14)
<SEAGATE ST9146803SS FS03>         at scbus1 target 41 lun 0 =
(da13,pass15)
<SEAGATE ST9146803SS FS03>         at scbus1 target 42 lun 0 =
(da14,pass16)


# camcontrol inq ses0
pass2: <IBM-ESXS SAS EXP BP 61A6> Fixed Enclosure Services SCSI-4 device=20=

pass2: 150.000MB/s transfers, Command Queueing Enabled

# camcontrol inq da8
pass10: <ATA Samsung SSD 840 BB0Q> Fixed Direct Access SCSI-6 device=20
pass10: Serial Number S1D9NEADA08547X    =20
pass10: 150.000MB/s transfers, Command Queueing Enabled

But I've noticed that disk transfers to one of the SSDs peak around 200 =
MB/s, while they reach 400 MB/s with mps and
a LSI2008 card replacing the Invader.

Looking for the mrsas driver to download I stumbled upon this:
=
http://www.lsi.com/downloads/Public/RAID%20Controllers/RAID%20Controllers%=
20Common%20Files/6.602.01.00_MR_Free-BSD_Driver.txt

Bug Fixes/Enhancements: =20
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
SCGCQ00523785 (CSET) - Data Corruption observed on JBODs while running =
IOs with 32k block size.
SCGCQ00459075 (DFCT) - On converting from UG to JBOD and vice versa, fw =
sees reply msQOverflow and hit montask.


I'll try to install this driver on my 10-STABLE and see how it goes.





Borja.




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?779F5D6D-BCA2-4376-84BC-6B08A8F357D4>