RE: MySQL Performance Woody Vs Sarge
Late last year, I also found problems with MySQL 4 running on the Sarge 2.6
Kernels. The 2.4 Kernel on Sarge for the packaged MySQL was much faster, I
did not looked into why this is, since Sarge was not ready for release and
this was a test machine. On a second test machine, I did build MySQL from
source on the 2.6 Kernel and did not have the same performance problems as
the pre-packaged MySQL. Again, no explanation as to why.... FYI, this
database about 2GB with over a billion records split between the two primary
transaction tables.
Tim
-----Original Message-----
From: Simon [mailto:dev@networker.co.nz]
Sent: Wednesday, June 01, 2005 2:03 AM
To: debian-user@lists.debian.org
Subject: Re: MySQL Performance Woody Vs Sarge
Jacob S wrote:
>>>What does 'hdparm' return for the RAID drive on each server when you
>>>don't give it any options (ie. 'hdparm /dev/hda')? You might also
>>>include the same output for hdparm run each of the hard drives used
>>>to make the array.
>>>
>>>What are the specs on the hard drives/RAID setup on both servers?
>>>Are they hardware raid or software raid?
>>
>>Both software RAID 0, both use Promise 2 channel PCI cards, Hard
>>Drives are IBM deskstar(?) 7200 40GB.
>>live: # hdparm /dev/md0
>
> ^^^^^
> Woody
> faster
>>/dev/md0:
>> readonly = 0 (off)
>> readahead = 256 (on)
>> geometry = 25248/2/4, sectors = 80418048, start = 0
>>dev: hdparm /dev/md0
>
> ^^^^
> Sarge
> slower
>>/dev/md0:
>> BLKROGET failed: Invalid argument
>> geometry = 23632/2/4, sectors = 80405120, start = 0
Jacob - Nearly Right:
woody(dev) = faster (Linux version 2.4.18-bf2.4):
# hdparm /dev/md0
/dev/md0:
BLKROGET failed: Invalid argument
geometry = 23632/2/4, sectors = 80405120, start = 0
The above raid consists of:
# hdparm /dev/hde
/dev/hde:
multcount = 0 (off)
I/O support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 5005/255/63, sectors = 80418240, start = 0
busstate = 1 (on)
# hdparm /dev/hdg
/dev/hdg:
multcount = 0 (off)
I/O support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
nowerr = 0 (off)
readonly = 0 (off)
readahead = 8 (on)
geometry = 79780/16/63, sectors = 80418240, start = 0
busstate = 1 (on)
sarge(production) = slower (Linux version 2.6.8-2-k7):
# hdparm /dev/md0
/dev/md0:
readonly = 0 (off)
readahead = 256 (on)
geometry = 25248/2/4, sectors = 80418048, start = 0
The above raid consists of:
# hdparm /dev/hde
/dev/hde:
multcount = 0 (off)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 65535/16/63, sectors = 80418240, start = 0
# hdparm /dev/hdg
/dev/hdg:
multcount = 0 (off)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 65535/16/63, sectors = 80418240, start = 0
> [Cc'ed back to the debian-user list, as I'm sure there are people that
> know more about this than I do. :-)]
>
> What is the kernel version on each server? Are they both using the same
> kernel module for the Promise controller? I'm getting the two boxes
> confused... let me know if my notes above are correct.
See above.
> The md0 devices should consist of two /dev/hd* devices. Please show
> hdparm output for them as well.
See above.
> I'm grasping at straws a little bit, as the results aren't what I first
> expected (unless I'm mixing previous facts about the servers). But maybe
> somebody else here will have some more clues.
Thanks
Simon
--
To UNSUBSCRIBE, email to debian-user-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact
listmaster@lists.debian.org
Reply to: