SCSI Tape drive on PowerEdge 2970 / LSI2032

John LLOYD jal at mdacorporation.com
Fri Feb 26 18:45:27 CST 2010


> Date: Thu, 25 Feb 2010 13:59:57 -0000
> From: "Jon Marston" <jonm at januahortorum.com>
> Subject: SCSI Tape drive on PowerEdge 2970 / LSI2032
> To: <linux-poweredge at dell.com>
> Message-ID:
> 	<4B8CACC472FA764B8AA7D1832EC71F3C01F3D0EA at NEPTUNE.jhl.local>
> Content-Type: text/plain;	charset="us-ascii"
> 
> Hi,
> I don't think this is *quite* within the usual scope of this mailing
> list. But if it doesn't start working someone will get murdered at
> random; I assumed you guys might be up for preventing murder, at least.
> 
> I have a new PowerEdge 2970 which we purchased fitted with an LSI 2032
> SCSI card with rear connector.
> 
> This is connected to a Quantum LTO3 tape drive in a rack-mount jobbie,
> configured as SCSI ID 6. We have a conversion cable connecting the
> PE2970's VHDCI to the Quantum's LVD HD-68 connector. We have a
> terminator in the remaining HD-68 port on the Quantum.
> 
> The problem we're having is that the SCSI card identified the Drive at
> ID 6 with a name of 'PP at JPPH PHPRHPH 2', which doesn't sound right to
> me.
> The server is running CentOS 5.4 x86_64.
> 
> The relevant dmesg output is as follows:
> 
> 
> Fusion MPT base driver 3.04.07rh
> Fusion MPT SPI Host driver 3.04.07rh
> mptbase: ioc0: Initiating bringup
> mptscsih: ioc0: attempting target reset! (sc=ffff81047f58ecc0)
> mptscsih: ioc0: target reset: SUCCESS (sc=ffff81047f58ecc0)
> mptbase: ioc0: LogInfo(0x110a0000): F/W: Device Problem
> mptscsih: ioc0: attempting bus reset! (sc=ffff81047f58ecc0)
> mptscsih: ioc0: bus reset: SUCCESS (sc=ffff81047f58ecc0)
> mptscsih: ioc0: attempting host reset! (sc=ffff81047f58ecc0)
> mptbase: ioc0: Initiating recovery
> mptscsih: ioc0: host reset: SUCCESS (sc=ffff81047f58ecc0)
> Fusion MPT misc device (ioctl) driver 3.04.07rh
> mptctl: Registered with Fusion MPT base driver
> mptctl: /dev/mptctl @ (major,minor=10,220)
> 
> I take it the above output is not good? There is no 'st' device node
> which I would expect from a SCSI tape drive.
> 
> What am I doing wrong?
> 
> Thanks,
> 
> Jon Marston
> Janua Hortorum

The SCSI card may also be set to target 6, in which case the tape drive and the SCSI card are stepping on each other.  Change the target number of the drive.

Or the cable is broken or pins are broken.

If termination is bad, "mtst -f /dev/nst0 status" should still work, even if tar etc don't.

Finally, a power cycle of drive and computer never hurts.  Merely rebooting sometimes is not enough.


--John




More information about the Linux-PowerEdge mailing list