experiences with 3Ware 6400 3W-6400?

experiences with 3Ware 6400 3W-6400?

[Rev. Jeffrey Paul]

I'm thinking of buying one of these for a production server that I'm
setting up.  Has anyone had any experiences with these on linux, good or
bad?

-j

--
--------------------------------------------------------
 Rev. Jeffrey Paul    -datavibe-     sneak@datavibe.net
   aim:x736e65616b   pgp:0x15FA257E   phone:8777483467
    70E0 B896 D5F3 8BF4 4BEE 2CCF EF2F BA28 15FA 257E
--------------------------------------------------------

Re: experiences with 3Ware 6400 3W-6400?

[Joshua Baker-LePain]

On Thu, 26 Jun 2003 at 4:57pm, Rev. Jeffrey Paul wrote

> I'm thinking of buying one of these for a production server that I'm
> setting up.  Has anyone had any experiences with these on linux, good or
> bad?

As an IDE controller, it'll work great.  The hardware RAID performance 
(especially RAID 5) ain't gonna' be hot, though.  The 7500 series performs 
much better in hardware RAID.

-- 
Joshua Baker-LePain
Department of Biomedical Engineering
Duke University

Re: experiences with 3Ware 6400 3W-6400?

[David Rees]

Rev. Jeffrey Paul said:
>
> I'm thinking of buying one of these for a production server that I'm
> setting up.  Has anyone had any experiences with these on linux, good or
> bad?

I've been using a couple of them in either RAID 1 or RAID 10 configs for a
while now with good results.  Nothing bad has happenned, but also haven't
experienced any drive failures either.

-Dave

Re: experiences with 3Ware 6400 3W-6400?

[Maurice Hilarius]

With regards to your message at 04:17 PM 6/26/03, David Rees. Where you stated:
>Rev. Jeffrey Paul said:
> >
> > I'm thinking of buying one of these for a production server that I'm
> > setting up.  Has anyone had any experiences with these on linux, good or
> > bad?
>
>I've been using a couple of them in either RAID 1 or RAID 10 configs for a
>while now with good results.  Nothing bad has happenned, but also haven't
>experienced any drive failures either.
>
>-Dave


Why would you buy a 2 year old, nearly unsupported card?
?Heck I can sell you 8 port 3W-7810 cards for under $300
And they work really well in software RAID.



With our best regards,

Maurice W. Hilarius       Telephone: 01-780-456-9771
Hard Data Ltd.               FAX:       01-780-456-9772
11060 - 166 Avenue        mailto:maurice@harddata.com
Edmonton, AB, Canada      http://www.harddata.com/
    T5X 1Y3

Re: experiences with 3Ware 6400 3W-6400?

[David Rees]

Maurice Hilarius said:
>>Rev. Jeffrey Paul said:
>> >
>> > I'm thinking of buying one of these for a production server that I'm
>> > setting up.  Has anyone had any experiences with these on linux, good
>> > or bad?
>
> Why would you buy a 2 year old, nearly unsupported card?
> ?Heck I can sell you 8 port 3W-7810 cards for under $300
> And they work really well in software RAID.

I was going to mention the same thing in my post, but forgot.  I didn't
think that you could buy the 6400 new anymore, so where are you getting
them from?  The 7500 runs about $250 which is a bit expensive for real
low-budget servers.

-Dave

Re: experiences with 3Ware 6400 3W-6400?

[Corey McGuire]

http://castle.pricewatch.com/search/search.idq?qc=%223WARE%22*%20AND%20%227810%22*%20AND%20%40totalcost%3E0%20AND%20%40minorder=1&cr=3ware%207810&ne=13032&l=12986

But don't get the one from weirdstuff... that ones MINE! ;-)

*********** REPLY SEPARATOR  ***********

On 6/26/2003 at 4:57 PM Rev. Jeffrey Paul wrote:

>I'm thinking of buying one of these for a production server that I'm
>setting up.  Has anyone had any experiences with these on linux, good or
>bad?
>
>-j
>
>--
>--------------------------------------------------------
> Rev. Jeffrey Paul    -datavibe-     sneak@datavibe.net
>   aim:x736e65616b   pgp:0x15FA257E   phone:8777483467
>    70E0 B896 D5F3 8BF4 4BEE 2CCF EF2F BA28 15FA 257E
>--------------------------------------------------------
>-
>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




/\/\/\/\/\/\ Nothing is foolproof to a talented fool. /\/\/\/\/\/\

coreyfro@coreyfro.com
http://www.coreyfro.com/
http://stats.distributed.net/rc5-64/psearch.php3?st=coreyfro
ICQ : 3168059

-----BEGIN GEEK CODE BLOCK-----
GCS !d--(+) s: a- C++++$ UL++>++++ P+ L++>++++ E- W+++$ N++ o? K? w++++$>+++++$ O---- !M--- V- PS+++ PE++(--) Y+ PGP- t--- 5(+) !X- R(+) !tv b-(+) Dl++(++++) D++ G++(-) e>+++ h++(---) r++>+$ y++**>$ H++++ n---(----) p? !au w+ v- 3+>++ j- G'''' B--- u+++*** f* Quake++++>+++++$
------END GEEK CODE BLOCK------

Home of Geek Code - http://www.geekcode.com/
The Geek Code Decoder Page - http://www.ebb.org/ungeek//

Re: experiences with 3Ware 6400 3W-6400?

[Rev. Jeffrey Paul]

On Fri, 27 Jun 2003, David Rees wrote:

> Maurice Hilarius said:
> >>Rev. Jeffrey Paul said:
> >> >
> >> > I'm thinking of buying one of these for a production server that I'm
> >> > setting up.  Has anyone had any experiences with these on linux, good
> >> > or bad?
> >
> > Why would you buy a 2 year old, nearly unsupported card?
> > ?Heck I can sell you 8 port 3W-7810 cards for under $300
> > And they work really well in software RAID.
>
> I was going to mention the same thing in my post, but forgot.  I didn't
> think that you could buy the 6400 new anymore, so where are you getting
> them from?  The 7500 runs about $250 which is a bit expensive for real
> low-budget servers.
>
> -Dave

I'd done some quick searches for "3ware" on ebay, having heard good things
about their ide-raid cards.  The 7xxx series are too expensive for the
project I'm working on at the moment.  It's either an older card, or
software RAID1 (and, with all due respect, I don't want to use linux
software raid in -this sort- of production environment).

Are the 6xxx series really that bad?  I don't care about 'nearly'
unsupported, just as long as it works.  My budget for this whole project
is likely less than $500.

We're currently looking at raid1 across two 80-120GB disks for a samba
file server in an office that does a lot of graphic design.

-j

--
--------------------------------------------------------
 Rev. Jeffrey Paul    -datavibe-     sneak@datavibe.net
   aim:x736e65616b   pgp:0x15FA257E   phone:8777483467
    70E0 B896 D5F3 8BF4 4BEE 2CCF EF2F BA28 15FA 257E
--------------------------------------------------------

Re: experiences with 3Ware 6400 3W-6400?

[Gordon Henderson]

On Fri, 27 Jun 2003, Rev. Jeffrey Paul wrote:

> I'd done some quick searches for "3ware" on ebay, having heard good things
> about their ide-raid cards.  The 7xxx series are too expensive for the
> project I'm working on at the moment.  It's either an older card, or
> software RAID1 (and, with all due respect, I don't want to use linux
> software raid in -this sort- of production environment).

Er.. Why not?

> We're currently looking at raid1 across two 80-120GB disks for a samba
> file server in an office that does a lot of graphic design.

I've setup several small servers doing just this with 2 x IDE drives
mirrored (RAID1) using the on-board controllers on the motherboard
concerned. Make both drives masters, connect them up, pick your Linux
distribution and off you go..

There is the problem that the dual on-board IDE controllers are usually
inside the same physical chip, and if that fails you'll lose both drives
until you replace the mobo, but you are in the same boat if a plug-in PCI
RAID card fails too. I've not (yet?) had problems with a failed drive
stopping the 2nd drive from working, as long as they are on separate
controllers (ie 2 masters off the mobo controller) I have seen one failed
drive take out a 2nd on the same bus though, so thats not recomended.
(once the failed drive was removed, the other did work correctly though)

Eg. the PC I'm currently typing this email on is 100 miles away and I
don't particularly want to drive to it to fix it if a disk fails...

gordon @ unicorn: df -h /
Filesystem            Size  Used Avail Use% Mounted on
/dev/md0              236M   20M  204M   9% /

gordon @ unicorn: cat /proc/mdstat
...
md0 : active raid1 hdc1[1] hda1[0]
      249856 blocks [2/2] [UU]
...

I've been using Linux software RAID1 on this particular machine for well
over 2 years now. This is one of my own personal servers and 2.5 years ago
I had to think hard about spending the extra on a 2nd drive. These days I
wouldn't think twice about doing it.

Gordon

Re: experiences with 3Ware 6400 3W-6400?

[Mike Dresser]

On Thu, 26 Jun 2003, Joshua Baker-LePain wrote:

> As an IDE controller, it'll work great.  The hardware RAID performance
> (especially RAID 5) ain't gonna' be hot, though.  The 7500 series performs

Now that's an understatement :)

P4 2.53, Asus P4B533-e, 3ware 7500-4, 4 Maxtor Maxline 250 gb drives in
raid5.

mickey,512M,17840,60,19878,5,8683,2,28247,89,91356,8,243.0,0,16,2576,81,+++++,+++,+++++,+++,3184,96,+++++,+++,9392,96

P4 2.53, Asus P4B533, 3ware 7500-4, 2 Maxtor DiamondMax 9 200 gb drives in
Raid1. (different machine)

r1,512M,26616,94,52362,15,21394,5,28918,92,47633,4,301.8,0,16,3145,99,+++++,+++,+++++,+++,3214,99,+++++,+++,10229,95

Half to a third the write speed, although the read speed is a lot
better.

Haven't upgraded to the 7.6 firmware yet(still using 7.5.3 that was on
the card), anyone notice any difference?

The 7000-2 and 7500-4's in raid1 work quite nicely.  The 7500-4 in raid5
actually lags the system for 3-5 seconds when doing a lot of disk i/o.
Feels like working on a p120 with DMA shut off, to be honest.

Works for what I'm doing with it though.

Mike

Re: experiences with 3Ware 6400 3W-6400?

[Lars Marowsky-Bree]

On 2003-06-27T14:43:23,
   Gordon Henderson <gordon@drogon.net> said:

> > project I'm working on at the moment.  It's either an older card, or
> > software RAID1 (and, with all due respect, I don't want to use linux
> > software raid in -this sort- of production environment).
> Er.. Why not?

Because the software RAID doesn't take care of problems like hotswap for
one.

That's about the only compelling reason for specialized hardware I can
really think of, but it's a good one.


Sincerely,
    Lars Marowsky-Brée <lmb@suse.de>

-- 
SuSE Labs - Research & Development, SuSE Linux AG
  
"If anything can go wrong, it will." "Chance favors the prepared (mind)."
  -- Capt. Edward A. Murphy            -- Louis Pasteur
-
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: experiences with 3Ware 6400 3W-6400?

[Maurice Hilarius]

With regards to your message at 09:53 AM 6/27/03, Lars Marowsky-Bree. Where 
you stated:
>On 2003-06-27T09:50:44,
>    Maurice Hilarius <maurice@harddata.com> said:
>
> > If you look at 3Wares site, you will find that they have their CLI 
> utility,
> > which supports hot-swap capability.
>
>Uhm, yes, exactly.
>
>I'm a 3ware user myself. ;-) I was saying that a strict software-raid
>solution was missing this.


Pardon?

If you run software RAID and use their CLI you have hot-swap.
If you run their BIOS controlled RAID and use their 3DM you get similar 
functions.

About the only difference I see is that 3DM has a pretty interface GUI in 
your web browser.

If that is important one could make a similar thing using tools like Python 
or Zope pretty easily..



With our best regards,

Maurice W. Hilarius       Telephone: 01-780-456-9771
Hard Data Ltd.               FAX:       01-780-456-9772
11060 - 166 Avenue        mailto:maurice@harddata.com
Edmonton, AB, Canada      http://www.harddata.com/
    T5X 1Y3

Re: experiences with 3Ware 6400 3W-6400?

[David Rees]

Rev. Jeffrey Paul said:
>
> I'd done some quick searches for "3ware" on ebay, having heard good things
> about their ide-raid cards.  The 7xxx series are too expensive for the
> project I'm working on at the moment.  It's either an older card, or
> software RAID1 (and, with all due respect, I don't want to use linux
> software raid in -this sort- of production environment).
>
> Are the 6xxx series really that bad?  I don't care about 'nearly'
> unsupported, just as long as it works.  My budget for this whole project
> is likely less than $500.
>
> We're currently looking at raid1 across two 80-120GB disks for a samba
> file server in an office that does a lot of graphic design.

For that type of usage, the 6xxx series will work out just fine.  There's
nothing wrong with them in RAID 1, 0 or 10.  Just their RAID 5 performance
is a bit lacking.

Looks like you can find them for about $60 these days so they're a good
deal.  Although you can find a 72xx card for about $15-20 more as well.

-Dave

Re: experiences with 3Ware 6400 3W-6400?

[David Rees]

Gordon Henderson said:
>
> Eg. the PC I'm currently typing this email on is 100 miles away and I
> don't particularly want to drive to it to fix it if a disk fails...
>
> gordon @ unicorn: df -h /
> Filesystem            Size  Used Avail Use% Mounted on
> /dev/md0              236M   20M  204M   9% /
>
> gordon @ unicorn: cat /proc/mdstat
> ...
> md0 : active raid1 hdc1[1] hda1[0]
>       249856 blocks [2/2] [UU]
> ...

So what do you do when /dev/hda fails and the computer won't reboot?  Have
you managed to get it to boot off /dev/hdc when /dev/hda isn't available?

Not having to worry about these sorts of things is one reason it's nice to
have hardware RAID for at least the boot device.

-Dave

Re: experiences with 3Ware 6400 3W-6400?

[Rev. Jeffrey Paul]

On Fri, 27 Jun 2003, David Rees wrote:

> Gordon Henderson said:
> >
> > Eg. the PC I'm currently typing this email on is 100 miles away and I
> > don't particularly want to drive to it to fix it if a disk fails...
> >
> > gordon @ unicorn: df -h /
> > Filesystem            Size  Used Avail Use% Mounted on
> > /dev/md0              236M   20M  204M   9% /
> >
> > gordon @ unicorn: cat /proc/mdstat
> > ...
> > md0 : active raid1 hdc1[1] hda1[0]
> >       249856 blocks [2/2] [UU]
> > ...
>
> So what do you do when /dev/hda fails and the computer won't reboot?  Have
> you managed to get it to boot off /dev/hdc when /dev/hda isn't available?
>
> Not having to worry about these sorts of things is one reason it's nice to
> have hardware RAID for at least the boot device.

Exactly.  Another issue is that I want them to be able to just swap out
(not necessarily hot, but offline even) a failed drive without my
intervention.... I assume the card will just resync it?  Under linux, it'd
require mucking about with mdadm or the raidtools and I want this to be a
fire-and-forget kind of system.  On solid hardware on a UPS, the only
things likely to die are the processor fan, the power supply, and the
drives...  all of which can be replaced with no knowledge of the raid or
software configuration.

Additionally, I didn't realize the 7xxx cards were that cheap.  If i'm
just using raid 1 or possibly raid10, is it really going to make that much
of a difference?

-j

--
--------------------------------------------------------
 Rev. Jeffrey Paul    -datavibe-     sneak@datavibe.net
   aim:x736e65616b   pgp:0x15FA257E   phone:8777483467
    70E0 B896 D5F3 8BF4 4BEE 2CCF EF2F BA28 15FA 257E
--------------------------------------------------------

Re: experiences with 3Ware 6400 3W-6400?

[Gordon Henderson]

On Fri, 27 Jun 2003, David Rees wrote:

> Gordon Henderson said:
> >
> > Eg. the PC I'm currently typing this email on is 100 miles away and I
> > don't particularly want to drive to it to fix it if a disk fails...
> >
> > gordon @ unicorn: df -h /
> > Filesystem            Size  Used Avail Use% Mounted on
> > /dev/md0              236M   20M  204M   9% /
> >
> > gordon @ unicorn: cat /proc/mdstat
> > ...
> > md0 : active raid1 hdc1[1] hda1[0]
> >       249856 blocks [2/2] [UU]
> > ...
>
> So what do you do when /dev/hda fails and the computer won't reboot?  Have
> you managed to get it to boot off /dev/hdc when /dev/hda isn't available?

Yes. Works a treat. No point doing stuff if you don't test it enough to
trust it.

You do need a modern mobo/BIOS combination that will boot off something
other than the first IDE drive, but thats been fairly standard for a few
years now IIRC.

I've recently built another system with 5 IDE drives; hde,g,i,k,m and it
boots off e which is mirrored on i and that boots also when e isn't
present.

> Not having to worry about these sorts of things is one reason it's nice to
> have hardware RAID for at least the boot device.

You do have a valid point there though, especially on older system that
can't do this. Maybe SCSI too - I haven't had a chance to build an
all-SCSI system for some time, but I get a chance in a few weeks time, so
that'll be nice to experiment with.

Gordon

Re[2]: experiences with 3Ware 6400 3W-6400?

[Alex Verbitsky]

Hello Rev. Jeffrey Paul,

Friday, June 27, 2003, 10:45:08 PM, you wrote:
RJP> Exactly.  Another issue is that I want them to be able to just swap out
RJP> (not necessarily hot, but offline even) a failed drive without my
RJP> intervention.... I assume the card will just resync it?  Under linux, it'd
RJP> require mucking about with mdadm or the raidtools and I want this to be a
RJP> fire-and-forget kind of system.  On solid hardware on a UPS, the only
RJP> things likely to die are the processor fan, the power supply, and the
RJP> drives...  all of which can be replaced with no knowledge of the raid or
RJP> software configuration.
you are wrong that 3ware card automatically resync replaced drive. I've
done some research with 7500-8 and it showed that:
1) 3ware card doesn't detect that drive was inserted - you should
   manually add it to the list using 3DM or CLI
2) after you added the drive you should _manually_ start rebuilding
   the array with that drive again using 3DM or CLI
3) OR you have to go to 3ware BIOS on boot - then new drive will be
   detected automatically and you just have to start rebuilding with ths
   drive. BUT:
      - after reboot rebuild will continue on first array access
        (mounting volume or running fdisk) not on start-up
      - if you reboot after drive swap and do nothing in BIOS you will
        not be able to use newly inserted drive for array rebuild
        since BIOS adds it as independed 1-disk JBOD array.
        Until 7.6 3ware software you had to go to 3ware BIOS in this
        situation. With 7.6 software you can handle this since add and
        delete array commands are now available.

I am not 100% sure about 1) since i did not have a chance to test it with
3ware drive cage (which might have some logic for signaling drive
inesrtion to the card), but on the hand i looked through 3ware manual and
did not see any mention of auto-detection - only manual operations.

If you think i'm wrong or have any info on this I am all ears.

I believe that is possible to create some set of scripts for
monitoring 3ware array and auto-resyncng replaced drive but that would
be limited to single array on card i think.
Actually i am now trying to write script of this kind using perl and CLI.

-- 
Best regards,
 Alex                            mailto:alexver5@mail.ru

Re: 3ware 7500 performance, was: experiences with 3Ware 6400 3W-6400?

[Mikael Abrahamsson]

On Sat, 28 Jun 2003, Alex Verbitsky wrote:

> What do you mean by this? I did not notice anything like this, though i
> use it for sequental i/o mainly. if your i/o is highly random, then
> 3ware acceleration (R5 fusion) is not working and you should receive
> something like 5-10 MB/s.

Which is exactly what we all get in real-life file server environment when
using onboard RAID5 without a lot of onboard cache (though I have seen the
same behaviour even on SX6000 with 128 meg ram). Filesystem read is
phenomenally fast (I have seen 150 meg+ off of 8 disk arrays) but write is 
very slow (5-10 meg/s).  Sequential write is also fast, but that never 
happens in a fileserver environment.

Since I use my raid5 mainly for long time cheap storage I don't have a big 
problem with all this, stability and ease of management is what I'm 
looking for, so I choose hardware raid5 on 3w7500.

-- 
Mikael Abrahamsson    email: swmike@swm.pp.se

3ware 7500 performance, was: experiences with 3Ware 6400 3W-6400?

[Alex Verbitsky]

Hello Mike,

Friday, June 27, 2003, 5:36:09 PM, you wrote:
MD> P4 2.53, Asus P4B533-e, 3ware 7500-4, 4 Maxtor Maxline 250 gb drives in
MD> raid5.
MD> mickey,512M,17840,60,19878,5,8683,2,28247,89,91356,8,243.0,0,16,2576,81,+++++,+++,+++++,+++,3184,96,+++++,+++,9392,96

MD> P4 2.53, Asus P4B533, 3ware 7500-4, 2 Maxtor DiamondMax 9 200 gb drives in
MD> Raid1. (different machine)
MD> r1,512M,26616,94,52362,15,21394,5,28918,92,47633,4,301.8,0,16,3145,99,+++++,+++,+++++,+++,3214,99,+++++,+++,10229,95

MD> Half to a third the write speed, although the read speed is a lot
MD> better.
What kernel did you use for testing?
I believe it makes difference: when i tried RH9 (2.4.20 kernel) i got
for 3 drive harware RAID5 - 11 MB/s seq write,
after downgrading to RH8 (2.4.18-14) i got 25-26 MB/s.
drives were ibm AVVA 80GB 7200RPM

when i added another 3 drives (ibm dtla, maxtor and barracuda 4 -
all 7200rpm) i got on 6 drive RAID5 - 29.5 MB/s (though CPU usage was
high - 50%)

upgrading to 2.4.21 made no difference (against 2.4.18),
upgrading firmware to 7.6 - the same.

my configuraton: Via Epia-M with Via C3-933mhz CPU, 256MB RAM, 3ware 7500-8

under Win2k on the same configuration i got 40-42 MB/s seq write for
4 drive harware RAID5 with 6% CPU usage.

As a conclusion, seems that RH9's kernel (and maybe some other kernels?) are broken.

Also, did you tweaked max-readahead/min-readahead params?
doing this:
        echo 256 > /proc/sys/vm/max-readahead
        echo 128 > /proc/sys/vm/min-readahead
increased my sequental read from 58 MB/s to 80 MB/s (at expense of CPU
usage - from 22% to 36%).

Also you did not report your RAM size. if you have 512 MB, your
results are not very accurate because some amount of writes is
cached.
I believe that test file should be at least twice size of the system
RAM, 4-6 times are better.

MD> Haven't upgraded to the 7.6 firmware yet(still using 7.5.3 that was on
MD> the card), anyone notice any difference?
I highly recommend you to upgrade - though i did not notice any
performance gains, it contains new improved CLI utility with online
array create/delete commands, and also better support RH8.

MD> The 7000-2 and 7500-4's in raid1 work quite nicely.  The 7500-4 in raid5
MD> actually lags the system for 3-5 seconds when doing a lot of disk i/o.
MD> Feels like working on a p120 with DMA shut off, to be honest.
What do you mean by this? I did not notice anything like this, though i
use it for sequental i/o mainly.
if your i/o is highly random, then 3ware acceleration (R5 fusion) is
not working and you should receive something like 5-10 MB/s.


-- 
Best regards,
 Alex                            mailto:alexver5@mail.ru

Re: 3ware 7500 performance, was: experiences with 3Ware 6400 3W-6400?

[Mike Dresser]

On Sat, 28 Jun 2003, Alex Verbitsky wrote:

> What kernel did you use for testing?

Those two machines are running 2.4.21, and saw similar results under
2.4.20.

> upgrading firmware to 7.6 - the same.

I'll try this next week when I'm at the office, and see if it does
anything.

> As a conclusion, seems that RH9's kernel (and maybe some other kernels?) are broken.

Using a self built kernel, so who knows.

> Also, did you tweaked max-readahead/min-readahead params?
> doing this:
>         echo 256 > /proc/sys/vm/max-readahead
>         echo 128 > /proc/sys/vm/min-readahead
> increased my sequental read from 58 MB/s to 80 MB/s (at expense of CPU
> usage - from 22% to 36%).

Haven't touched those, testing that now.

Yup, seeing about a 10-15% increase for the most part, and a little more
cpu load, which is't a problem, as I've got spare cpu to spare, even with
the gzip going.  Can't very accurately test the system as I can't just
init 1 it and go test from here.

> Also you did not report your RAM size. if you have 512 MB, your
> results are not very accurate because some amount of writes is
> cached.

256 meg in each machine.

> I highly recommend you to upgrade - though i did not notice any
> performance gains, it contains new improved CLI utility with online
> array create/delete commands, and also better support RH8.

I have one machine running 7.6 already, and didn't notice anything useful
to me, besides being able to call up the alarms from the cli, which I
track in the syslog anyways.

If I have a drive failure, I have to power down the machine anyways, so
online create/delete doesn't bother me much.

> MD> The 7000-2 and 7500-4's in raid1 work quite nicely.  The 7500-4 in raid5
> MD> actually lags the system for 3-5 seconds when doing a lot of disk i/o.
> MD> Feels like working on a p120 with DMA shut off, to be honest.
> What do you mean by this? I did not notice anything like this, though i
> use it for sequental i/o mainly.
> if your i/o is highly random, then 3ware acceleration (R5 fusion) is
> not working and you should receive something like 5-10 MB/s.

Likely that, as if I'm ungzip'ing say a 3 gig file, and then go to log in,
doing even an ls will take a few seconds to come up.

Mike

Re: 3ware 7500 performance, was: experiences with 3Ware 6400 3W-6400?

[Mike Dresser]

On Sat, 28 Jun 2003, Mikael Abrahamsson wrote:

> Since I use my raid5 mainly for long time cheap storage I don't have a big
> problem with all this, stability and ease of management is what I'm
> looking for, so I choose hardware raid5 on 3w7500.

Exactly, that's what I'm using this machine for.  And the ones that are
raid1 don't show this behaviour.

Mike

Re[2]: 3ware 7500 performance, was: experiences with 3Ware 6400 3W-6400?

[Alex Verbitsky]

Hello Mikael,

Saturday, June 28, 2003, 4:42:58 PM, you wrote:
MA> Which is exactly what we all get in real-life file server environment when
MA> using onboard RAID5 without a lot of onboard cache (though I have seen the
MA> same behaviour even on SX6000 with 128 meg ram). Filesystem read is
MA> phenomenally fast (I have seen 150 meg+ off of 8 disk arrays) but write is 
MA> very slow (5-10 meg/s).  Sequential write is also fast, but that never 
MA> happens in a fileserver environment.
Exactly. One seriously looking for good random write performance should use
RAID10 - it's nearly as fast as RAID0 on 3ware. But that's another
story - it's not so cheap.

BTW re RAID performance i can recommend this excelent testing
http://www.fcenter.ru/articles.shtml?hdd/2216#4

He tested 3w7810 but I believe for random access 7810 and 7500-8 is
equivalent. It is in Russian, but all significant info (diagrams, charts) in
English.

-- 
Best regards,
 Alex                            mailto:alexver5@mail.ru