Compatible Hardware Controllers (Smartctl, 8 port, SFF-8087)

Compatible Hardware Controllers (Smartctl, 8 port, SFF-8087)

[Andrew Dunn]

I am running two decent sized mdadm arrays:

####################################################

storrgie@MINERVA:~$ cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5]
[raid4] [raid10]
md2 : active raid6 sdi1[4] sdl1[7] sdf1[1] sdk1[6] sdj1[5] sdh1[3]
sdg1[2] sde1[0]
      11721071616 blocks level 6, 512k chunk, algorithm 2 [8/8] [UUUUUUUU]
      [===>.................]  check = 18.8% (368163248/1953511936)
finish=504.2min speed=52401K/sec

md0 : active raid6 sdn1[4] sdt1[6] sdm1[3] sdr1[7] sdq1[0] sds1[2]
sdo1[1] sdd1[5]
      5860558848 blocks level 6, 512k chunk, algorithm 2 [8/8] [UUUUUUUU]
      [==============>......]  check = 71.2% (696254448/976759808)
finish=87.8min speed=53242K/sec

md1 : active raid0 sdc1[1] sdb1[0]
      586067072 blocks 64k chunks

unused devices: <none>

####################################################

These two arrays are on this controller:

http://www.supermicro.com/products/accessories/addon/AOC-USAS-L8i.cfm
http://blog.agdunn.net/?p=391

The problem with this controller is that if you try to use smartctl or
smartd to monitor your drives, the controller will become unresponsive
and offline an arbitrary amount of drives. This causes a huge problem
in raid because then the drives get marked as failed and I have to
force add them back into the array.

My solution is basically to move from raid5 to raid6 so that I can
afford a single drive failure, because I am not preempting a drive
malfunction.

I would very much like to know if any of you are running budget
($100-300) controller cards that work well with both mdadm and
smartctl. I would very much like the card to support SFF-8087
connectors but they don't have to.

Thank you all in advance.

Re: Compatible Hardware Controllers (Smartctl, 8 port, SFF-8087)

[Drew]

> The problem with this controller is that if you try to use smartctl or
> smartd to monitor your drives, the controller will become unresponsive
> and offline an arbitrary amount of drives. This causes a huge problem
> in raid because then the drives get marked as failed and I have to
> force add them back into the array.

This is a known issue with the 1068 series which AFAIK there are some
patches floating around that resolve this issue. Refer to
https://bugzilla.kernel.org/show_bug.cgi?id=14831 for particulars.


-- 
Drew

"Nothing in life is to be feared. It is only to be understood."
--Marie Curie

Re: Compatible Hardware Controllers (Smartctl, 8 port, SFF-8087)

[Cláudio Martins]

On Tue, 4 May 2010 12:09:41 -0700 Drew <drew.kay@gmail.com> wrote:
> > The problem with this controller is that if you try to use smartctl or
> > smartd to monitor your drives, the controller will become unresponsive
> > and offline an arbitrary amount of drives. This causes a huge problem
> > in raid because then the drives get marked as failed and I have to
> > force add them back into the array.
> 
> This is a known issue with the 1068 series which AFAIK there are some
> patches floating around that resolve this issue. Refer to
> https://bugzilla.kernel.org/show_bug.cgi?id=14831 for particulars.
> 

 Hi,

 Please take a look at

 https://bugzilla.kernel.org/show_bug.cgi?id=13594

 As it seems to refer to the same issue.

 A patch is referred on the bottom of the bug page, but I haven't tried
it yet.


On Tue, 4 May 2010 14:45:29 -0400 Andrew Dunn <andrew.g.dunn.dod@gmail.com> wrote:
> My solution is basically to move from raid5 to raid6 so that I can
> afford a single drive failure, because I am not preempting a drive
> malfunction.

 Do note that this probably won't do you any good on the long run. I
have this issue and I've seen it kick out several drives in a row, so
you can end up with a dead array even it it's raid6.

 Please let me know if you have any luck with any of the pacthes, and
I'll do so also.

 Best regards.

Cláudio

--
To unsubscribe from this list: send the line "unsubscribe linux-raid" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html