mps questions

[Borja Marcos]

Hello,

I''m trying some Dell PERC H200 cards on FreeBSD 8.2 with v28 ZFS. (Not a production environment, just running tests).

I have two variants of them. One is an internal one, identified as this one.

mps0: <LSI SAS2008> port 0xfc00-0xfcff mem 0xdf2b0000-0xdf2bffff,0xdf2c0000-0xdf2fffff irq 32 at device 0.0 on pci2
mps0: Firmware: 02.15.63.00
mps0: IOCCapabilities: 185c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,IR>
mps0: [ITHREAD]


And another one intended to conect to external SAS devices.

mps1: <LSI SAS2008> port 0xec00-0xecff mem 0xdf4b0000-0xdf4bffff,0xdf4c0000-0xdf4fffff irq 41 at device 0.0 on pci5
mps1: Firmware: 02.15.63.00
mps1: IOCCapabilities: 185c<ScsiTaskFull,DiagTrace,SnapBuf,EEDP,TransRetry,IR>
mps1: [ITHREAD]

Both have two SAS ports. And I am trying them on Dell PowerEdge R510 servers. I have 8 internal hard disks attached, and a Dell PowerVault MD1220 connected to one of the external SAS connectors. The MD1220 has 8 hard disks as well.

I haven't configured any logical volumes on the cards, I just want unobtrusive SAS cards in order to create ZFS pools on raw disks, and the setup is working great, except for one problem:

The *internal* card is showing just six hard disks under FreeBSD, not eight.

da0 at mps0 bus 0 scbus0 target 1 lun 0
da1 at mps0 bus 0 scbus0 target 2 lun 0
da2 at mps0 bus 0 scbus0 target 3 lun 0
da3 at mps0 bus 0 scbus0 target 4 lun 0
da4 at mps0 bus 0 scbus0 target 5 lun 0
da5 at mps0 bus 0 scbus0 target 6 lun 0

And this happens no matter how I connect the internal SAS connectors. There are two connectors from the disk backplanes to the card, SAS0 and SAS1 (I assume they are two buses) and even if I reverse them I still see eight disks. 

However, when booting the system I see eight disks. The controller  BIOS lists the backplane device and the eight disks, the FreeBSD boot loader says it sees eight disks, but FreeBSD sees just six.

# camcontrol rescan all
Re-scan of bus 0 was successful
Re-scan of bus 1 was successful
# camcontrol devlist
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 1 lun 0 (da0,pass0)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 2 lun 0 (da1,pass1)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 3 lun 0 (da2,pass2)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 4 lun 0 (da3,pass3)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 5 lun 0 (da4,pass4)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 6 lun 0 (da5,pass5)
<Alcor Flash Disk 8.07>            at scbus1 target 0 lun 0 (da6,pass6)

The card with the outside facing connectors shows the eight hard disks atttached to the MD1120 and the disks. A camcontrol devlist on the machine with both mps cards installed I see this:

# camcontrol devlist
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 1 lun 0 (da0,pass0)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 2 lun 0 (da1,pass1)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 3 lun 0 (da2,pass2)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 4 lun 0 (da3,pass3)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 5 lun 0 (da4,pass4)
<SEAGATE ST3300657SS-H EH02>       at scbus0 target 6 lun 0 (da5,pass5)
<SEAGATE ST9146852SS HT64>         at scbus1 target 1 lun 0 (da6,pass6)
<SEAGATE ST9146852SS HT64>         at scbus1 target 2 lun 0 (da7,pass7)
<SEAGATE ST9146852SS HT64>         at scbus1 target 3 lun 0 (da8,pass8)
<SEAGATE ST9146852SS HT64>         at scbus1 target 4 lun 0 (da9,pass9)
<SEAGATE ST9146852SS HT64>         at scbus1 target 5 lun 0 (da10,pass10)
<SEAGATE ST9146852SS HT64>         at scbus1 target 6 lun 0 (da11,pass11)
<SEAGATE ST9146852SS HT64>         at scbus1 target 7 lun 0 (da12,pass12)
<SEAGATE ST9146852SS HT64>         at scbus1 target 13 lun 0 (da13,pass13)
<DELL MD1220 1.01>                 at scbus1 target 36 lun 0 (ses0,pass14)
<UT165 USB Flash Disk 0.00>        at scbus2 target 0 lun 0 (pass15,da14)


Any ideas? I was wondering if Dell had done something nasty to the internal card firmware (mps0 in my example), but seems that the eight internal disks are correctly presented to the BIOS.




Thank you,





Borja.