Marvel AOC-SASLP-MV8 - 8 Sata PCI-E 4x non-raid controller card

Hi simonsez,

it works !! :cool:

I was able to compile AND load the driver on ubuntu server 10.04 with the default kernel (2.6.32-21-generic-pae #32-Ubuntu SMP).

lsmod:

mvsas 69394 0
libsas 45086 1 mvsas
scst 290308 1 mvsas
scsi_transport_sas 26289 2 mvsas,libsas

dmesg:

[ 165.907252] scst: ***WARNING***: Patch io_context was not applied on your kernel. SCST will be working with not the best performance.
[ 165.918083] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918172] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918221] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918259] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918630] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918840] [792]: scst: scst_register_device:792:Attached to scsi0, channel 0, id 0, lun 0, type 5
[ 165.918910] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918917] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918923] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918929] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918941] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918949] [792]: scst: scst_register_device:792:Attached to scsi2, channel 0, id 0, lun 0, type 0
[ 165.918957] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918962] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918968] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918973] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918981] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918987] [792]: scst: scst_register_device:792:Attached to scsi2, channel 0, id 1, lun 0, type 0
[ 165.919583] [793]: scst: scst_cmd_thread:3729:
[ 165.919634] [794]: scst: scst_cmd_thread:3729:processing thread started, PID 793
[ 165.919668] Processing thread started, PID 794
[ 165.920626] [795]: scst: scst_init_thread:3482:Init thread started, PID 795
[ 165.921113] [796]: scst: scst_tm_thread:5092:Task management thread started, PID 796
[ 165.922283] [792]: scst: init_scst:2089:SCST version 2.0.0-pre1-procfs loaded successfully (max mem for commands 217MB, per device 86MB)
[ 165.922370] [792]: scst: scst_print_config:1859:Enabled features: EXTRACHECKS, DEBUG
[ 165.923383] [797]: scst: scst_global_mgmt_thread:5663:Management thread started, PID 797

Next i´ll try to add the patches to get rid of the warning in the first line. Unfortunately i´m not able to test the driver with a real workload, because i had to rma my memory. Hopefully the new ones will arrive this weekend.

Big thanks for uploading that scst source for me (and the others) :D

greetings
 
Last edited:
[Dodger];1035866822 said:
Hi simonsez,

it works !! :cool:

I was able to compile AND load the driver on ubuntu server 10.04 with the default kernel (2.6.32-21-generic-pae #32-Ubuntu SMP).

lsmod:

mvsas 69394 0
libsas 45086 1 mvsas
scst 290308 1 mvsas
scsi_transport_sas 26289 2 mvsas,libsas

dmesg:

[ 165.907252] scst: ***WARNING***: Patch io_context was not applied on your kernel. SCST will be working with not the best performance.
[ 165.918083] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918172] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918221] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918259] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918630] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918840] [792]: scst: scst_register_device:792:Attached to scsi0, channel 0, id 0, lun 0, type 5
[ 165.918910] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918917] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918923] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918929] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918941] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918949] [792]: scst: scst_register_device:792:Attached to scsi2, channel 0, id 0, lun 0, type 0
[ 165.918957] [792]: scst_suspend_activity:599:suspend_count 0
[ 165.918962] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918968] [792]: scst_suspend_activity:644:Waiting for 0 active commands finally to complete
[ 165.918973] [792]: scst_susp_wait:578:wait_event() returned 0
[ 165.918981] [792]: __scst_resume_activity:675:suspend_count 0 left
[ 165.918987] [792]: scst: scst_register_device:792:Attached to scsi2, channel 0, id 1, lun 0, type 0
[ 165.919583] [793]: scst: scst_cmd_thread:3729:
[ 165.919634] [794]: scst: scst_cmd_thread:3729:processing thread started, PID 793
[ 165.919668] Processing thread started, PID 794
[ 165.920626] [795]: scst: scst_init_thread:3482:Init thread started, PID 795
[ 165.921113] [796]: scst: scst_tm_thread:5092:Task management thread started, PID 796
[ 165.922283] [792]: scst: init_scst:2089:SCST version 2.0.0-pre1-procfs loaded successfully (max mem for commands 217MB, per device 86MB)
[ 165.922370] [792]: scst: scst_print_config:1859:Enabled features: EXTRACHECKS, DEBUG
[ 165.923383] [797]: scst: scst_global_mgmt_thread:5663:Management thread started, PID 797

Next i´ll try to add the patches to get rid of the warning in the first line. Unfortunately i´m not able to test the driver with a real workload, because i had to rma my memory. Hopefully the new ones will arrive this weekend.

Big thanks for uploading that scst source for me (and the others) :D

greetings
Click to expand...

Hey Dodger,

Damn! I was just gonna try it tonight and wanted to post my results back, but you beat me to it. Glad you got it working. :cool:

I am going to give it a go on 10.04 too, sounds like it's easy. Report back on how you're making out with it if you can when you get your new memory. For me running using Bonnie against my my raid 6 array was a sure fire way to put load on the driver and test it out. The non-working drivers always failed when running a bonnie test on the drive.

I wonder if this should be reported to the folks working on scst or the people working on the mvsas kernel driver? I'm not even sure it would help them, and I don't know how to go about doing that.

Si.
 
[Dodger];1035866822 said:
Hi simonsez,

it works !! :cool:

I was able to compile AND load the driver on ubuntu server 10.04 with the default kernel (2.6.32-21-generic-pae #32-Ubuntu SMP).
Click to expand...

Hey Doger,

Got around to doing my own test this weekend and I'm afraid this isn't going to work. :( I built a fresh 10.04 system, recompiled the old driver and adjusted. After booting everything seemed to look good:

Jun 27 14:41:59 xxxxxxxx kernel: [ 13.280727] sd 6:0:0:0: Attached scsi generic sg2 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.280881] sd 6:0:0:0: [sdc] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.281119] sd 6:0:0:0: [sdc] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.281119] sd 6:0:0:0: [sdc] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.281305] sd 6:0:1:0: Attached scsi generic sg3 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.281435] sd 6:0:0:0: [sdc] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.281989] sd 6:0:2:0: Attached scsi generic sg4 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.282039] sdc:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.282547] sd 6:0:3:0: Attached scsi generic sg5 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.283615] sd 6:0:4:0: Attached scsi generic sg6 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.290017] sd 6:0:5:0: Attached scsi generic sg7 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.290017] sd 6:0:6:0: Attached scsi generic sg8 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.290201] sd 6:0:7:0: Attached scsi generic sg9 type 0
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.291261] sd 6:0:1:0: [sdd] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.291261] sd 6:0:1:0: [sdd] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.291261] sd 6:0:1:0: [sdd] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292172] sd 6:0:1:0: [sdd] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292172] sdd: unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:0:0: [sdc] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:2:0: [sde] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:2:0: [sde] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:2:0: [sde] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:2:0: [sde] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sde:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:3:0: [sdf] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:3:0: [sdf] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:3:0: [sdf] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:3:0: [sdf] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:4:0: [sdg] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sdf:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:4:0: [sdg] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:4:0: [sdg] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:4:0: [sdg] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sdg:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:5:0: [sdh] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:5:0: [sdh] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:5:0: [sdh] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sd 6:0:5:0: [sdh] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.292505] sdh:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:6:0: [sdi] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:6:0: [sdi] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:6:0: [sdi] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:6:0: [sdi] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sdi:
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:7:0: [sdj] 3907029168 512-byte logical blocks: (2.00 TB/1.81 TiB)
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:7:0: [sdj] Write Protect is off
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:7:0: [sdj] Mode Sense: 00 3a 00 00
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sd 6:0:7:0: [sdj] Write cache: enabled, read cache: enabled, doesn't support DPO or FUA
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.293757] sdj: unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.299721] sd 6:0:1:0: [sdd] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.300786] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.301488] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.301630] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.302899] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.304461] sd 6:0:4:0: [sdg] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.304554] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.305023] sd 6:0:6:0: [sdi] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.305527] sd 6:0:3:0: [sdf] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.305813] sd 6:0:2:0: [sde] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.306385] sd 6:0:5:0: [sdh] Attached SCSI disk
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.309765] unknown partition table
Jun 27 14:41:59 xxxxxxxx kernel: [ 13.309765] sd 6:0:7:0: [sdj] Attached SCSI disk
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.393520] md: bind<sdc>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.395152] md: bind<sdh>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.402825] md: bind<sdg>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.408965] md: bind<sdj>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.411826] md: bind<sde>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.420168] md: bind<sdi>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.426846] md: bind<sdd>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.612311] md: bind<sdf>
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618657] raid5: device sdf operational as raid disk 3
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618657] raid5: device sdd operational as raid disk 1
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618657] raid5: device sdi operational as raid disk 6
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618667] raid5: device sde operational as raid disk 2
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618676] raid5: device sdj operational as raid disk 7
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618684] raid5: device sdg operational as raid disk 4
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618692] raid5: device sdh operational as raid disk 5
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.618701] raid5: device sdc operational as raid disk 0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621132] raid5: allocated 8490kB for md_d5
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 3: w=1 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 1: w=2 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 6: w=3 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 2: w=4 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 7: w=5 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 4: w=6 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 5: w=7 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] 0: w=8 pa=0 pr=8 m=2 a=2 r=8 op1=0 op2=0
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] raid5: raid level 6 set md_d5 active with 8 out of 8 devices, algorithm 2
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] RAID5 conf printout:
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] --- rd:8 wd:8
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 0, o:1, dev:sdc
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 1, o:1, dev:sdd
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 2, o:1, dev:sde
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 3, o:1, dev:sdf
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 4, o:1, dev:sdg
un 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 5, o:1, dev:sdh
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 6, o:1, dev:sdi
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] disk 7, o:1, dev:sdj
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] md_d5: detected capacity change from 0 to 12002393063424
Jun 27 14:42:00 xxxxxxxx mdadm[1195]: NewArray event detected on md device /dev/md/d5
Jun 27 14:42:00 xxxxxxxx kernel: [ 13.621176] md_d5: unknown partition table
Jun 27 14:42:03 xxxxxxxx kernel: [ 16.480077] eth0: no IPv6 routers present
Jun 27 14:42:11 xxxxxxxx ntpd[1049]: peer 172.27.100.1 now valid
Jun 27 14:44:29 xxxxxxxx kernel: [ 163.118597] JFS: nTxBlock = 8192, nTxLock = 65536
Click to expand...

So I went ahead and mounted all the drive. Things looked good, and I could see all the data on my array. Next test was to run bonnie to put some load.

When I did I got the same drive dropouts:

Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010285] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1669:mvs_abort_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010296] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1608:mvs_query_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010331] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1669:mvs_abort_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010342] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1608:mvs_query_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010374] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1669:mvs_abort_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010385] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1608:mvs_query_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010412] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1669:mvs_abort_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010423] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1608:mvs_query_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010453] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1669:mvs_abort_task:rc= 5
Jun 27 14:48:11 xxxxxxxx kernel: [ 385.010464] /build/buildd/linux-2.6.32/drivers/scsi/mvsas/mv_sas.c 1608:mvs_query_task:rc= 5
Click to expand...

I don't think this is the actual drive dropping out, it's a problem between the driver and the card. After a short while the drives will start to be ejected from the array. At this point I did a shutdown and went back to my 9.10, and everything is working again. Because I couldn't shutdown cleanly I had to do a jfs_fsck on my drive and now mdadm wants to resync the whole array. A process that'll easily take 15 hours.

Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md_d5 : active raid6 sdj[7] sdi[6] sdh[5] sdg[4] sdf[3] sde[2] sdd[1] sdc[0]
11721086976 blocks level 6, 1024k chunk, algorithm 2 [8/8] [UUUUUUUU]
[>....................] resync = 0.9% (17931672/1953514496) finish=923.0min speed=34947K/sec
Click to expand...

So it looks like I'm stuck on 9.10 for now. I am going to try some other ideas, but when this happens I like to let the whole array sync before testing again.

Anyways, I'd like a second opinion so if you get a chance to test with drives I'd like to know if you get the same thing. As I said, bonnie is an excellent test for this. In my working config it works just fine, under the test config bonnie will cause the problem a few seconds after being run.

Cheers.

Si.
 
Hi simonsez,

i had exactly the same issues. BUT *drumroll* i´m able to access the drives with the lates mainline kernel.

2.6.35-020635rc3-generic

Though i get the following errors in dmesg:

[ 9.386309] EXT4-fs (sda1): re-mounted. Opts: errors=remount-ro
[ 9.561389] ata10: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.561448] ata10: status=0x01 { Error }
[ 9.561451] ata10: error=0x04 { DriveStatusError }
[ 9.561460] ata9: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.561513] ata9: status=0x01 { Error }
[ 9.561515] ata9: error=0x04 { DriveStatusError }
[ 9.561523] ata13: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.561577] ata13: status=0x01 { Error }
[ 9.561579] ata13: error=0x04 { DriveStatusError }
[ 9.561588] ata12: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.561641] ata12: status=0x01 { Error }
[ 9.561643] ata12: error=0x04 { DriveStatusError }
[ 9.561652] ata11: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.561706] ata11: status=0x01 { Error }
[ 9.561707] ata11: error=0x04 { DriveStatusError }
[ 9.564834] ata11: translated ATA stat/err 0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.564889] ata11: status=0x51 { DriveReady SeekComplete Error }
[ 9.564892] ata11: error=0x04 { DriveStatusError }
[ 9.565116] ata11: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565193] ata11: status=0x01 { Error }
[ 9.565195] ata11: error=0x04 { DriveStatusError }
[ 9.565200] ata10: translated ATA stat/err 0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565277] ata10: status=0x51 { DriveReady SeekComplete Error }
[ 9.565279] ata10: error=0x04 { DriveStatusError }
[ 9.565285] ata13: translated ATA stat/err 0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565361] ata13: status=0x51 { DriveReady SeekComplete Error }
[ 9.565364] ata13: error=0x04 { DriveStatusError }
[ 9.565370] ata12: translated ATA stat/err 0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565446] ata12: status=0x51 { DriveReady SeekComplete Error }
[ 9.565449] ata12: error=0x04 { DriveStatusError }
[ 9.565467] ata9: translated ATA stat/err 0x51/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565543] ata9: status=0x51 { DriveReady SeekComplete Error }
[ 9.565545] ata9: error=0x04 { DriveStatusError }
[ 9.565680] ata13: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565749] ata13: status=0x01 { Error }
[ 9.565751] ata13: error=0x04 { DriveStatusError }
[ 9.565772] ata10: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565826] ata10: status=0x01 { Error }
[ 9.565828] ata10: error=0x04 { DriveStatusError }
[ 9.565837] ata12: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565891] ata12: status=0x01 { Error }
[ 9.565893] ata12: error=0x04 { DriveStatusError }
[ 9.565915] ata9: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[ 9.565968] ata9: status=0x01 { Error }
[ 9.565970] ata9: error=0x04 { DriveStatusError }

But the drives stay operational. Yesterday i built my raid5 with 5 drives and it synced successfully. I also did some tests with hdparm and bonnie++.

While i was formatting the raid i had one reset of a drive. But the formatting just stopped for a few seconds and continued then. No corruption on the raid so far.

The following benchmarks are on a crypted drive, so the speeds look very slow compared to yours.

bonnie++

1.96,1.96,xxxx,1,1277745138,8G,,640,95,108531,13,44785,14,3104,84,168566,22,546.8,12,16,,,,,21270,35,+++++,+++,+++++,+++,27109,39,+++++,+++,+++++,+++,18674us,19 13ms,9045ms,51104us,111ms,839ms,358us,1181us,468us,609us,63us,40us

hdparm -tT /dev/md0 (uncrypted)
/dev/md0:
Timing cached reads: 3982 MB in 2.00 seconds = 1991.68 MB/sec
Timing buffered disk reads: 1212 MB in 3.00 seconds = 403.74 MB/sec

same crypted:
/dev/mapper/raid5_0:
Timing cached reads: 4010 MB in 2.00 seconds = 2004.72 MB/sec
Timing buffered disk reads: 368 MB in 3.01 seconds = 122.30 MB/sec

in the end some dd writing:

10240+0 in
10240+0 out
10737418240 Bytes (11 GB) kopiert, 93,3147 s, 115 MB/s

Throughout the whole testing no new error messages in the log.

So at least we got some kind of a "stable" driver in the kernel. The question is is it stable enough to put valuable data on the drives.
 
srengr said:
An update. I have finally given up on the AOC-SASLP-MV8 cards. Over the weekend, my WHS server implementation decided that several drives were suddenly "missing'. The system was basically idling at the time with no real I/O going on. I was able to get WHS to resync by pulling and re-racking each drive in sequence. I then copied all the data on the array to a backup drive (which took all weekend).

I installed the Atto card this morning. WHS booted and is currently updating the array. I'll report on how the Atto card performs over time.

SM really needs to work out the bugs with their new card. It is really disappointing since every other product I have used from them has been first rate.

--srengr
Click to expand...

I think I am just about over this card also. Every time I try to add the last drive another drive ends up going missing during any type of heavy load whether it is cleaning up backups or actually doing the backup.

It seems like I spend more time trying to fix my WHS then use it.
 
Hmmm, looks like I'm not the only one who is "missing" drives periodically. In my case I just shut down my WHS and start it back within minnute or two. Magically "missing" drives are back online and everything is good. I'm sure it's SASLP-MV8 card playing games with me. I don't see any other explanation.
 
The only problem I have with my WHS and these cards are if I tell it to shutdown, it turns back on a second after it shuts off, Doesn't do that with cards removed, but seeing as I leave it on 24/7 its never usually a problem.
 
Hi,

yes cartman123. The module which is included in the 2.6.35-020635rc3-generic kernel. I think it will be supported in later kernels too.

After i have copied some terabytes i can say the driver seems stable enough.

Bye, Dodg
 
That is very good news indeed. With 2.6.34 I can only use jfs, as btrfs and xfs result in the card locking the system(I don't use RAID). Furthermore, when using jfs, if I try to copy files via thunar (a graphical file manager), the card locks up, but when I use cp on bash there is no problem.
 
Can anyone confirm that these do *not* work in Windows XP x86? I have read the whole thread and it seems not a single person has even tried? In fact it looks like no one has even tried to use it in XP x64 which is basically just Windows 2003. It probably would work with XP x64 even if it is officially unsupported. I would definitely like to try one of these cards, but if it doesn't have drivers that work with XP x86 I think it may be worth another $50+ to get a card that does (like LSI or Highpoint).
 
Is anyone still working on this?

I ordered three of these cards for 2 mdadm machines running debian lenny but I couldnt get the cards to work properly. The cards were properly detected but the drives werent. A kernel update to the latest experimental fixed this, but now the drives still drop randomly and hotswapping results in a kernel panic.

I have tried compiling new kernels, with or without patches. I've tried the scst mvsas.ko driver compiles with no good results. After reading that some people have the card working properly under ubuntu 9.10 it got my hopes up. But after starting to build the raid6 array I was quickly dissapointed and saw the drives getting marked as faulty spares one by one.

I have contacted Supermicro support, their first reply was really fast (2 hours) but after my reply stating my system spec it went quiet.

If anyone has this card working under any linux distribution can you reply with the kernel, driver version and distro you are using?

If there is no way to fix this problem, is there any other controllers you guys can recommend?
 
Hi Woutern83,

i use Linux xxx 2.6.35-020635rc3-generic #020635rc3 SMP Sat Jun 12 09:09:12 UTC 2010 x86_64 GNU/Linux as kernel in Ubuntu Server 10.04.

02:00.0 SCSI storage controller: Marvell Technology Group Ltd. MV64460/64461/64462 System Controller, Revision B (rev 01)

It works more or less good in a raid 5 with 5 x 2TB Western Digital drives.

/dev/md0:
Version : 00.90
Creation Time : Sat Jun 26 13:15:32 2010
Raid Level : raid5
Array Size : 7814057984 (7452.07 GiB 8001.60 GB)
Used Dev Size : 1953514496 (1863.02 GiB 2000.40 GB)
Raid Devices : 5
Total Devices : 5
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Sun Aug 22 08:50:41 2010
State : active
Active Devices : 5
Working Devices : 5
Failed Devices : 0
Spare Devices : 0

Layout : left-symmetric
Chunk Size : 128K

UUID : c0636231:f6201db8:e8def76a:ab29c0ef ()
Events : 0.65

Number Major Minor RaidDevice State
0 8 16 0 active sync /dev/sdb
1 8 32 1 active sync /dev/sdc
2 8 48 2 active sync /dev/sdd
3 8 64 3 active sync /dev/sde
4 8 80 4 active sync /dev/sdf

From time to time i get some weird messages in syslog:

[1343603.010090] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 1701:<7>mv_abort_task() mvi=ffff88021a540000 task=ffff88012b87c8c0 slot=ffff88021a5665d0 slot_idx=x2
[1343603.010106] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 1632:mvs_query_task:rc= 5
[1343603.010155] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2081:port 3 ctrl sts=0x89800.
[1343603.010165] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2083:port 3 irq sts = 0x1001001
[1343603.010178] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2109:phy3 Unplug Notice
[1343603.020217] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2081:port 3 ctrl sts=0x199800.
[1343603.020223] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2083:port 3 irq sts = 0x1001081
[1343603.023869] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2081:port 3 ctrl sts=0x199800.
[1343603.023879] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2083:port 3 irq sts = 0x10000
[1343603.023888] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 2136:notify plug in on phy[3]
[1343603.033842] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 1224:port 3 attach dev info is 0
[1343603.033842] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 1226:port 3 attach sas addr is 3
[1343603.033842] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 378:phy 3 byte dmaded.
[1343605.230130] /home/kernel-ppa/COD/linux/drivers/scsi/mvsas/mv_sas.c 1586:mvs_I_T_nexus_reset for device[3]:rc= 0
[1343605.230150] ata12: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[1343605.235258] ata12: status=0x01 { Error }
[1343605.235265] ata12: error=0x04 { DriveStatusError }

But the raid stays operational. it seemes the driver has some timing issues.
 
That's good news. I just installed 2.6.35 on my system, seems that the mv_abort_task errors and subsequent controller freezes that i got when i was copying files with thunar have disappeared. Now I am trying all the other things that froze the controller, and will report back.

PS. When i am booting my fileserver, I get Port 3 errors from the MV8, and i need to press any key to continue. Everything works ok after that, but thats a PITA for a headless server(That only happens when I have the HP SAS expander connected)
 
Thanks for the reply [Dodger],

I'm having some sort of progress with Ubuntu 9.10 Karmic and 2.6.36-rc1. The array stays in tact, but becomes unresponsive quite often.
I have noticed that a mdadm --detail /dev/md0 immediatly unfreezes the array. If I just wait it can last up to 15-20 seconds, but as soon as I noticed a hang and do mdadm --detail /dev/md0 the array unfreezes.

Code: 
[11276.970038] drivers/scsi/mvsas/mv_sas.c 2024:phy3 Attached Device
[11277.010080] drivers/scsi/mvsas/mv_sas.c 2083:port 3 ctrl sts=0x89800.
[11277.010090] drivers/scsi/mvsas/mv_sas.c 2085:Port 3 irq sts = 0x1001
[11277.010100] drivers/scsi/mvsas/mv_sas.c 2111:phy3 Unplug Notice
[11277.010136] drivers/scsi/mvsas/mv_sas.c 2083:port 3 ctrl sts=0x199800.
[11277.010143] drivers/scsi/mvsas/mv_sas.c 2085:Port 3 irq sts = 0x81
[11277.014300] drivers/scsi/mvsas/mv_sas.c 2083:port 3 ctrl sts=0x199800.
[11277.014307] drivers/scsi/mvsas/mv_sas.c 2085:Port 3 irq sts = 0x10000
[11277.014314] drivers/scsi/mvsas/mv_sas.c 2138:notify plug in on phy[3]
[11277.024292] drivers/scsi/mvsas/mv_sas.c 1224:port 3 attach dev info is 0
[11277.024292] drivers/scsi/mvsas/mv_sas.c 1226:port 3 attach sas addr is 3
[11277.024292] drivers/scsi/mvsas/mv_sas.c 378:phy 3 byte dmaded.
[11277.180073] drivers/scsi/mvsas/mv_sas.c 1586:mvs_I_T_nexus_reset for device[3]:rc= 0
[11277.180087] ata10: translated ATA stat/err 0x01/04 to SCSI SK/ASC/ASCQ 0xb/00/00
[11277.180095] ata10.00: device reported invalid CHS sector 0
[11277.180101] ata10: status=0x01 { Error }
[11277.180107] ata10: error=0x04 { DriveStatusError }

It seems this combination still has problems with fake-drops. I'm happy the array is still intact. But writing or reading from the array gets interrupted all the time, causing me to be annoyed :p

I have 2 machines to test with (2x Norco 4220, 3 cards and 8+ disks each). On one im testing with Ubuntu 9.10. I'm testing the other one with Debian Lenny and 2.6.36-rc2 and I haven't seen any errors yet. So I'm carefully hopeful.
 
After 5 days of testing, I can attest that mvsas with 2.6.35 works fine, with no errors, freezes or anything like that. I have an HP SAS Expander + 13x1.5TB single drives with JFS and no RAID and my OS is Arch Linux

@Woutern83:
I don't use RAID, so I cant offer anything on that front, but I would suggest that you post the errors you get on the linux scsi mailing list
 
I'll give that a go cartman123, thanks for the suggestion.

I'll dismantle my array and test the disks seperatly, let's see what happens.
 
Hi everyone,

good news (i think). With the recent kernel 2.6.36-020636rc2-generic the problems seem to be solved.
I got no error messages in the log and no freezes of my raid anymore. I hope this will last. I don´t know what the kernel hackers have done, but it looks good. :D

I have followed the linux-scsi mailing list, but there is nothing mentioned that would have caused that change.

Greetings
 
There are some patches that seem to be relevant,
Jiri Slaby at 22/6, James Bottomley at 27/7, one from David Milburn at 6/8 and two from Dan Carpenter at 11/8
 
Any idea what it fixes? Running it in WHS so wondering it I need to upgrade the firmware.
 
With some of the issues I have with drives going missing sometimes I would hate to make it worse lol
 
brennok said:
With some of the issues I have with drives going missing sometimes I would hate to make it worse lol
Click to expand...

Same here. And it happened before i upgrade frimware. I just lost a driver in port number 2 . After a few time lost driver in window i have return my card and get a new one. everything is ok now. I think it may be because the MV8 card so hot, it hot like crazy when i touch in it heatsink. I am thinking of change the heatsink or place a small fan direct to it heatsink.
 
Yeh a fan blowing onto the card is a must... I have a 120 sunnon blowwing onto 2 and it keeps them nice and cool ;)
 
Does anyone actually know what the bios update is for?
I don't like to update firmwares if I dont know what they fix.
 
Just ordered 1 today and the back order time is only 4 weeks. i can't wait to test it. Price tag was around 145 USD (incl VAT).
 
Hi,

[Dodger] stated it seemed stable with the 2.6.36-rc2 kernel. Can anyone comment on this? Has anyone had it running for quite some time? We have a server with 3 of these cards that currently has the 4 SATA disks in it attached to the onboard SATA controller due to stability issues, but it would need expansion soon and there are no more onboard ports free.
 
Hi freakynl,

unfortunately i was wrong. I still have issues with 2.6.36-020636-generic and also with the latest next kernel. As far as i have followed the scsi mailing listthe problems seem to be solved. Just a matter of time when the patches will go into the next kernel.

greetings,
dodge
 
Thanks for the re' :).

That's quite unfortunate news. Any idea which patch on the mailinglist supposedly fixed it? There are quite a few :). Is it this one?

diff --git a/drivers/scsi/libsas/sas_ata.c b/drivers/scsi/libsas/sas_ata.c
index 042153c..ddbade7 100644
--- a/drivers/scsi/libsas/sas_ata.c
+++ b/drivers/scsi/libsas/sas_ata.c
@@ -347,6 +347,7 @@ static int sas_ata_scr_read(struct ata_link *link, unsigned int sc_reg_in,
static struct ata_port_operations sas_sata_ops = {
.phy_reset = sas_ata_phy_reset,
.post_internal_cmd = sas_ata_post_internal,
+ .qc_defer = ata_std_qc_defer,
.qc_prep = ata_noop_qc_prep,
.qc_issue = sas_ata_qc_issue,
.qc_fill_rtf = sas_ata_qc_fill_rtf,


TIA

[edit]Just checked and this patch indeed isn't part of 2.6.36 (vanilla)[/edit]
 
Hello,
I am looking for a card that allows me to have more than 4 hard drives in my system and I think the AOC-SASLP-MV8 meets my needs.
I like to know if you advise me this card to create a software RAID? (I think create a RAID5 with 5 drives in 2To on a Ubuntu 10.04 or this kind of linux)
I see that some people have problems following the linux kernel and I do not really want me to confront the same problems.

thank you for your reply and sorry for my poor English.

Sincerely,
Marcellusio

EDIT: I read all the pages and I come to the conclusion that we must wait for the new version but which one ?
Wait & See
 
Last edited:
For the life of me cant get this card to work in WHS. Someone have a link with more driver options? Right now I just get "Primary IDE Channel" wigh a yellow exclamation mark beside it and it won't let me use the drivers I got with the card.
 
I have been following this thread passively waiting for some advice on how to get it working on linux. From kernel 2.6.36 (also rc) it looked quite good but I still had timeouts (also still saw NCQ errors in the log, but not every error seemed to be related to a timeout) for example when streaming a movie, it recovered after 5 seconds but it was still annoying. I recently installed the kernel 2.6.37-rc3 and that seemed to solve the issues. It's now running for 2 weeks or so and I have not seen any glitches (neither any strange log entries). Perhaps now the patches freakynl describes above are in there ...

For reference:
debian stable + kernel 2.6.37-rc3 from kernel.org compiled myself (also there is an ipv6 bug in that one which i patched by hand so probably better to go for 2.6.37-rc4 which is out for a couple of days) running a linux soft raid 6 with 5x1.5T seagate disks.

As far as I read this thread ... I would go for a more expensive card and save myself the trouble of investing so much time and effort in getting this to work (but perhaps from now on it works at least for linux), vendors should get shot for releasing hardware which does not even work good on any stock windows/linux
 
errors are still not 100 percent solved it seems ... but not really noticed this one during the copy process (probably during a copy of ~ 500GB of files to my raid 6)

Dec 5 21:13:11 fileserver kernel: ata15.00: device reported invalid CHS sector 0
Dec 5 21:13:11 fileserver kernel: ata15: status=0x01 { Error }
Dec 5 21:13:11 fileserver kernel: ata15: error=0x04 { DriveStatusError }

currently running 2.6.37-rc4
 
Okay for those mistakes but what is the impact the raid volume?
I did not really feel that this affects performance.
Otherwise, someone knows the meaning of this error?
 
marcellusio said:
Okay for those mistakes but what is the impact the raid volume?
I did not really feel that this affects performance.
Otherwise, someone knows the meaning of this error?
Click to expand...

As I -perhaps not so explicitly- mentioned in the post:
I did not notice it at the time ... so my raid functions properly and the file transfer did not stop as far as I know ... performance could have been down any level at that moment but I did not notice it in any case ...
 
Konowl said:
For the life of me cant get this card to work in WHS. Someone have a link with more driver options? Right now I just get "Primary IDE Channel" wigh a yellow exclamation mark beside it and it won't let me use the drivers I got with the card.
Click to expand...

Anyone?
 
You must log in or register to reply here.
Back
Top