RAID 5 throughputs

RAID 5 throughputs

[jassduec]

Hi All,

I have a storage system with 12 Maxtor 7200 rpm 250GB hard drives in
RAID 5 configuration with one hot spare. Maxtor claims that each drive
is capable of providing 56 MB/s sustained throughput. My host system
is connected to it using 64bit ultra320 Adaptec SCSI card. The host
system has 3.0 GHz pentium 4 (with HT) processor and 4 GB of RAM
running linux kernel 2.6.9. I was wondering what order of "maximum"
sequential read/write, random read/write throughput should i expect
using them. I am using some of the benchmarking tools like Bonnie,
Bonnie++, tiobench, dd, iometer etc. However the results from them
have been very depressing. Theoretically i should be able to achieve a
read sequential throughput close to min(10*56 = 560, 320) = 320 MB/s.
However my results are no way close to it. I was wondering if anybody
can give me some insight of what order of practical throughputs i
should expect using linux and what other factors may be creating
bottlenecks.

TIA

Re: RAID 5 throughputs

[Justin Piszcz]

On Sat, 19 Aug 2006, jassduec@gmail.com wrote:

> Hi All,
>
> I have a storage system with 12 Maxtor 7200 rpm 250GB hard drives in
> RAID 5 configuration with one hot spare. Maxtor claims that each drive
> is capable of providing 56 MB/s sustained throughput. My host system
> is connected to it using 64bit ultra320 Adaptec SCSI card. The host
> system has 3.0 GHz pentium 4 (with HT) processor and 4 GB of RAM
> running linux kernel 2.6.9. I was wondering what order of "maximum"
> sequential read/write, random read/write throughput should i expect
> using them. I am using some of the benchmarking tools like Bonnie,
> Bonnie++, tiobench, dd, iometer etc. However the results from them
> have been very depressing. Theoretically i should be able to achieve a
> read sequential throughput close to min(10*56 = 560, 320) = 320 MB/s.
> However my results are no way close to it. I was wondering if anybody
> can give me some insight of what order of practical throughputs i
> should expect using linux and what other factors may be creating
> bottlenecks.
>
> TIA
> -
> To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>

Is your 64bit ultra320 card in a 64bit PCI + 66MHZ slot?

If not, you will be limited to 133mb/s.

Re: RAID 5 throughputs

[jassduec]

Yeah, infact it is in a PCI-X slot @66Mhz.

On 8/19/06, Justin Piszcz <jpiszcz@lucidpixels.com> wrote:
>
>
> On Sat, 19 Aug 2006, jassduec@gmail.com wrote:
>
> > Hi All,
> >
> > I have a storage system with 12 Maxtor 7200 rpm 250GB hard drives in
> > RAID 5 configuration with one hot spare. Maxtor claims that each drive
> > is capable of providing 56 MB/s sustained throughput. My host system
> > is connected to it using 64bit ultra320 Adaptec SCSI card. The host
> > system has 3.0 GHz pentium 4 (with HT) processor and 4 GB of RAM
> > running linux kernel 2.6.9. I was wondering what order of "maximum"
> > sequential read/write, random read/write throughput should i expect
> > using them. I am using some of the benchmarking tools like Bonnie,
> > Bonnie++, tiobench, dd, iometer etc. However the results from them
> > have been very depressing. Theoretically i should be able to achieve a
> > read sequential throughput close to min(10*56 = 560, 320) = 320 MB/s.
> > However my results are no way close to it. I was wondering if anybody
> > can give me some insight of what order of practical throughputs i
> > should expect using linux and what other factors may be creating
> > bottlenecks.
> >
> > TIA
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> >
>
> Is your 64bit ultra320 card in a 64bit PCI + 66MHZ slot?
>
> If not, you will be limited to 133mb/s.
>
>

Re: RAID 5 throughputs

[Justin Piszcz]

On Sat, 19 Aug 2006, jassduec@gmail.com wrote:

> Yeah, infact it is in a PCI-X slot @66Mhz.
>
> On 8/19/06, Justin Piszcz <jpiszcz@lucidpixels.com> wrote:
>> 
>> 
>> On Sat, 19 Aug 2006, jassduec@gmail.com wrote:
>> 
>> > Hi All,
>> >
>> > I have a storage system with 12 Maxtor 7200 rpm 250GB hard drives in
>> > RAID 5 configuration with one hot spare. Maxtor claims that each drive
>> > is capable of providing 56 MB/s sustained throughput. My host system
>> > is connected to it using 64bit ultra320 Adaptec SCSI card. The host
>> > system has 3.0 GHz pentium 4 (with HT) processor and 4 GB of RAM
>> > running linux kernel 2.6.9. I was wondering what order of "maximum"
>> > sequential read/write, random read/write throughput should i expect
>> > using them. I am using some of the benchmarking tools like Bonnie,
>> > Bonnie++, tiobench, dd, iometer etc. However the results from them
>> > have been very depressing. Theoretically i should be able to achieve a
>> > read sequential throughput close to min(10*56 = 560, 320) = 320 MB/s.
>> > However my results are no way close to it. I was wondering if anybody
>> > can give me some insight of what order of practical throughputs i
>> > should expect using linux and what other factors may be creating
>> > bottlenecks.
>> >
>> > TIA
>> > -
>> > To unsubscribe from this list: send the line "unsubscribe linux-scsi" in
>> > the body of a message to majordomo@vger.kernel.org
>> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
>> >
>> 
>> Is your 64bit ultra320 card in a 64bit PCI + 66MHZ slot?
>> 
>> If not, you will be limited to 133mb/s.
>> 
>> 
>

What speeds are you currently seeing, r/w?

Have you tried making a 3-drive raid5 and increasing it by 1 drive at a 
time to see if there is any difference in speed?

Justin.