Re: extremly low memory usage - Mailing list pgsql-performance

Ron wrote:
> At 02:53 PM 8/20/2005, Jeremiah Jahn wrote:
>
>> On Fri, 2005-08-19 at 16:03 -0500, John A Meinel wrote:
>> > Jeremiah Jahn wrote:
>> > > On Fri, 2005-08-19 at 12:18 -0500, John A Meinel wrote:
>> > >
>> <snip>
>> >
>> > > it's cached alright. I'm getting a read rate of about 150MB/sec. I
>> would
>> > > have thought is would be faster with my raid setup. I think I'm
>> going to
>> > > scrap the whole thing and get rid of LVM. I'll just do a straight
>> ext3
>> > > system. Maybe that will help. Still trying to get suggestions for a
>> > > stripe size.
>> > >

Well, since you can get a read of the RAID at 150MB/s, that means that
it is actual I/O speed. It may not be cached in RAM. Perhaps you could
try the same test, only using say 1G, which should be cached.

>> >
>> > I don't think 150MB/s is out of the realm for a 14 drive array.
>> > How fast is time dd if=/dev/zero of=testfile bs=8192 count=1000000
>> >
>> time dd if=/dev/zero of=testfile bs=8192 count=1000000
>> 1000000+0 records in
>> 1000000+0 records out
>>
>> real    1m24.248s
>> user    0m0.381s
>> sys     0m33.028s
>>
>>
>> > (That should create a 8GB file, which is too big to cache everything)
>> > And then how fast is:
>> > time dd if=testfile of=/dev/null bs=8192 count=1000000
>>
>> time dd if=testfile of=/dev/null bs=8192 count=1000000
>> 1000000+0 records in
>> 1000000+0 records out
>>
>> real    0m54.139s
>> user    0m0.326s
>> sys     0m8.916s
>>
>>
>> and on a second run:
>>
>> real    0m55.667s
>> user    0m0.341s
>> sys     0m9.013s
>>
>>
>> >
>> > That should give you a semi-decent way of measuring how fast the RAID
>> > system is, since it should be too big to cache in ram.
>>
>> about 150MB/Sec. Is there no better way to make this go faster...?

I'm actually curious about PCI bus saturation at this point. Old 32-bit
33MHz pci could only push 1Gbit = 100MB/s. Now, I'm guessing that this
is a higher performance system. But I'm really surprised that your write
speed is that close to your read speed. (100MB/s write, 150MB/s read).

>
> Assuming these are U320 15Krpm 147GB HDs, a RAID 10 array of 14 of them
> doing raw sequential IO like this should be capable of at
>  ~7*75MB/s= 525MB/s using Seagate Cheetah 15K.4's,  ~7*79MB/s= 553MB/s
> if using Fujitsu MAU's, and ~7*86MB/s= 602MB/s if using Maxtor Atlas 15K
> II's to devices external to the RAID array.

I know I thought these were SATA drives, over 2 controllers. I could be
completely wrong, though.

>
> _IF_ the controller setup is high powered enough to keep that kind of IO
> rate up.  This will require a controller or controllers providing dual
> channel U320 bandwidth externally and quad channel U320 bandwidth
> internally.  IOW, it needs a controller or controllers talking 64b
> 133MHz PCI-X, reasonably fast DSP/CPU units, and probably a decent sized
> IO buffer as well.
>
> AFAICT, the Dell PERC4 controllers use various flavors of the LSI Logic
> MegaRAID controllers.  What I don't know is which exact one yours is,
> nor do I know if it (or any of the MegaRAID controllers) are high
> powered enough.
>
> Talk to your HW supplier to make sure you have controllers adequate to
> your HD's.
>
> ...and yes, your average access time will be in the 5.5ms - 6ms range
> when doing a physical seek.
> Even with RAID, you want to minimize seeks and maximize sequential IO
> when accessing them.
> Best to not go to HD at all ;-)

Well, certainly, if you can get more into RAM, you're always better off.
For writing, a battery-backed write cache, and for reading lots of
system RAM.

>
> Hope this helps,
> Ron Peacetree
>

John
=:->