Re: PERC5 - MegaRaid-SAS problems..

Re: PERC5 - MegaRaid-SAS problems..

[Gordon Henderson]

For anyone who cares about my saga so-far ;-) ...

I got physical access to the unit this morning and setup the drives as 15
RAID-0 Logical drives and booted up Linux, and it then attached all the
drives in the usual way.

And I can see all 15 drives. So the down-side is that I can't use any sort
of SMART stuff over the bus (as pointed out by someone else).

Anyway, it's currently in a RAID-1 configuration (which I used for some
initial soaktests) and seems to be just fine:

Filesystem            Size  Used Avail Use% Mounted on
/dev/md9              6.8T   33M  6.7T   1% /mnt

And it's more than fast enough too, although it'll be a lot slower when I
rebuilt it in RAID-6 mode.

Version  1.03       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
bertha           2G           266336  99 203934  61           421850  40 872.9   1


md9 : active raid0 sdo[14] sdn[13] sdm[12] sdl[11] sdk[10] sdj[9] sdi[8]
sdh[7] sdg[6] sdf[5] sde[4] sdd[3] sdc[2] sdb[1] sda[0]
      7317748800 blocks 64k chunks

Cheers,

Gordon

RE: PERC5 - MegaRaid-SAS problems..

[Patrick_Boyd]

You can use either LSI or Dell utilities to monitor the SMART status of
the disks.

Patrick Boyd
Dell Storage Software Engineer
(512)728-3182
 

-----Original Message-----
From: linux-raid-owner@vger.kernel.org
[mailto:linux-raid-owner@vger.kernel.org] On Behalf Of Gordon Henderson
Sent: Tuesday, October 17, 2006 5:45 AM
To: linux-poweredge-Lists; linux-raid@vger.kernel.org
Subject: Re: PERC5 - MegaRaid-SAS problems..


For anyone who cares about my saga so-far ;-) ...

I got physical access to the unit this morning and setup the drives as
15 RAID-0 Logical drives and booted up Linux, and it then attached all
the drives in the usual way.

And I can see all 15 drives. So the down-side is that I can't use any
sort of SMART stuff over the bus (as pointed out by someone else).

Anyway, it's currently in a RAID-1 configuration (which I used for some
initial soaktests) and seems to be just fine:

Filesystem            Size  Used Avail Use% Mounted on
/dev/md9              6.8T   33M  6.7T   1% /mnt

And it's more than fast enough too, although it'll be a lot slower when
I rebuilt it in RAID-6 mode.

Version  1.03       ------Sequential Output------ --Sequential Input-
--Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block--
--Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP
/sec %CP
bertha           2G           266336  99 203934  61           421850  40
872.9   1


md9 : active raid0 sdo[14] sdn[13] sdm[12] sdl[11] sdk[10] sdj[9] sdi[8]
sdh[7] sdg[6] sdf[5] sde[4] sdd[3] sdc[2] sdb[1] sda[0]
      7317748800 blocks 64k chunks

Cheers,

Gordon
-
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

Re: PERC5 - MegaRaid-SAS problems..

[Andrew Moise]

On 10/17/06, Gordon Henderson <gordon@drogon.net> wrote:
> Anyway, it's currently in a RAID-1 configuration (which I used for some
> initial soaktests) and seems to be just fine:
>
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/md9              6.8T   33M  6.7T   1% /mnt

  Incidentally, what filesystem are you using?  We may be setting up a
similar situation soon, and I'm considering ext3 and XFS.  ext3
apparently has a low max fs size, though, and I've heard bad things
about XFS's behavior on unclean shutdowns...

Re: PERC5 - MegaRaid-SAS problems..

[Greg Dickie]

Never lost an XFS filesystem completely. Can't say the same about ext3.

just my 2 cents,
Greg

On Tue, 2006-10-17 at 13:58 -0400, Andrew Moise wrote:
> On 10/17/06, Gordon Henderson <gordon@drogon.net> wrote:
> > Anyway, it's currently in a RAID-1 configuration (which I used for some
> > initial soaktests) and seems to be just fine:
> >
> > Filesystem            Size  Used Avail Use% Mounted on
> > /dev/md9              6.8T   33M  6.7T   1% /mnt
> 
>   Incidentally, what filesystem are you using?  We may be setting up a
> similar situation soon, and I'm considering ext3 and XFS.  ext3
> apparently has a low max fs size, though, and I've heard bad things
> about XFS's behavior on unclean shutdowns...
> 
> _______________________________________________
> Linux-PowerEdge mailing list
> Linux-PowerEdge@dell.com
> http://lists.us.dell.com/mailman/listinfo/linux-poweredge
> Please read the FAQ at http://lists.us.dell.com/faq
-- 
Greg Dickie
just a guy
Maximum Throughput

Re: PERC5 - MegaRaid-SAS problems..

[Gordon Henderson]

On Tue, 17 Oct 2006, Greg Dickie wrote:

> Never lost an XFS filesystem completely. Can't say the same about ext3.

Whereas I have exactly the reverse of the problem... Never lost an ext2/3,
but had a few XFSs trashed when I played with it a couple of years ago...

My 2 euros,
Gordon

_______________________________________________
Linux-PowerEdge mailing list
Linux-PowerEdge@dell.com
http://lists.us.dell.com/mailman/listinfo/linux-poweredge
Please read the FAQ at http://lists.us.dell.com/faq

Re: PERC5 - MegaRaid-SAS problems..

[Gordon Henderson]

On Tue, 17 Oct 2006, Andrew Moise wrote:

> On 10/17/06, Gordon Henderson <gordon@drogon.net> wrote:
> > Anyway, it's currently in a RAID-1 configuration (which I used for some
> > initial soaktests) and seems to be just fine:
> >
> > Filesystem            Size  Used Avail Use% Mounted on
> > /dev/md9              6.8T   33M  6.7T   1% /mnt
>
>   Incidentally, what filesystem are you using?  We may be setting up a
> similar situation soon, and I'm considering ext3 and XFS.  ext3
> apparently has a low max fs size, though, and I've heard bad things
> about XFS's behavior on unclean shutdowns...

I'm using ext3, and it's now RAID-6, so a little smaller:

Filesystem            Size  Used Avail Use% Mounted on
/dev/md9              5.9T   33M  5.7T   1% /archive

And slower:

Version  1.03       ------Sequential Output------ --Sequential Input- --Random-
                    -Per Chr- --Block-- -Rewrite- -Per Chr- --Block-- --Seeks--
Machine        Size K/sec %CP K/sec %CP K/sec %CP K/sec %CP K/sec %CP  /sec %CP
bertha           2G           186290  96 111105  75           262779  71 602.0 2


I've looked at, and tried XFS (and JFS) and keep coming back to ext3,
which I have on many, many servers and have never had an issue with it.

I have had problems with XFS, but that was about 2 years ago, so things
might have improved by then.

This is the biggest server I've setup so-far, it beats my previous record
of 3TB (which I have on 2 identical servers halfway round the world from
each other and they've been doing well for several months now).

After your comment about ext3 max FS size, I had a bit of a oo-er
situation so had to do a bit of looking to make sure I was OK, and it
seems that 16TB is the current limit, so OK there for a while, at least...
(/usr/src/linux/documentation/filesystems/ext2.txt)

Gordon

Re: PERC5 - MegaRaid-SAS problems..

[Andrew Moise]

On 10/17/06, Gordon Henderson <gordon@drogon.net> wrote:
> I have had problems with XFS, but that was about 2 years ago, so things
> might have improved by then.

  Well, filling some random files with zeroes because of an unclean
shutdown is still defined as "correct" behavior in XFS.  That hasn't
changed that I've heard of.  Why they haven't implemented something
akin to "data=ordered" is a mystery to me.
  I guess I'm begging to start a filesystem flamewar at this point :-).

> After your comment about ext3 max FS size, I had a bit of a oo-er
> situation so had to do a bit of looking to make sure I was OK, and it
> seems that 16TB is the current limit, so OK there for a while, at least...
> (/usr/src/linux/documentation/filesystems/ext2.txt)

  Aah, great news.  Different sources differ on what the max size is,
so I was pessimistically assuming it might be as low as 2TB.  Thanks.

Re: PERC5 - MegaRaid-SAS problems..

[Chris Allen]

Andrew Moise wrote:
> On 10/17/06, Gordon Henderson <gordon@drogon.net> wrote:
>> I have had problems with XFS, but that was about 2 years ago, so things
>> might have improved by then.
>
>  Well, filling some random files with zeroes because of an unclean
> shutdown is still defined as "correct" behavior in XFS.  That hasn't
> changed that I've heard of.  Why they haven't implemented something
> akin to "data=ordered" is a mystery to me.
>  I guess I'm begging to start a filesystem flamewar at this point :-).

I can consistently crash XFS over md (kernel panic) using a simple 
forking perl
script that copies large numbers of files about. The response from the 
XFS mailing
list was a shrug and a "well don't do that then..". Problem is, this 
exactly what
the clients of my machines *will* do. I can't crash ext3 in this way no 
matter
what I do.

>
>> After your comment about ext3 max FS size, I had a bit of a oo-er
>> situation so had to do a bit of looking to make sure I was OK, and it
>> seems that 16TB is the current limit, so OK there for a while, at 
>> least...
>> (/usr/src/linux/documentation/filesystems/ext2.txt)
>
>  Aah, great news.  Different sources differ on what the max size is,
> so I was pessimistically assuming it might be as low as 2TB.  Thanks.
> -

The max in mainstream kernels is 8TB. mm kernels will support 16TB.