Folks:
Thanks for the feedback. I've spent a good bit of time doing some testing. This
is what I've found. I'm also hoping to answer some of the questions that were
asked of me.
The network throughput I'm betting between machines that are hooked up to the
Gigabit switch were what I thought I expected - ~25MB/s. When I hooked one of
the machines up through the net so that the speed of the net was 100MB, this
speed dropped to 8MB/s. From what we can tell, these are realistic transfer
rates.
All eight disks are on the same controller. RAID/5.
When I WROTE from a client (MacOS X 10.6.2) to an NFS mounted drive, I got a
transfer rate of 3.1MB/s. When I READ from that same drive, I got a transfer
speed of 8MB/s.
While writing iostat gave the following data:
avg-cpu: %user %nice %system %iowait %steal %idle
1.19 0.00 1.68 23.85 0.00 73.28
Device: tps Blk_read/s Blk_wrtn/s Blk_read Blk_wrtn
sda 0.00 0.00 0.00 0 0
sda1 0.00 0.00 0.00 0 0
sda2 0.00 0.00 0.00 0 0
sda3 0.00 0.00 0.00 0 0
sda4 0.00 0.00 0.00 0 0
sda5 0.00 0.00 0.00 0 0
sdb 56.00 0.00 6921.60 0 34608
sdb1 56.00 0.00 6921.60 0 34608
Obviously the sdb device is what I was writing to. The Blk_wrtn/s comes out to
~3.2MB/s. I'm not sure how to interpret the %iowait statistic.
I'm a small place so I don't have any managed switches. They're auto sensing.
The ethernet ports on the computer in question are integral to the motherboard.
Since, during straight copies the transfer rate is the same (both transmit and
receive seem to give me about 8MB/s over a 100MB line I would tend to think the
I don't have a duplex mismatch.
I have replaced the network cables and tried different ports on the switch.
There is no difference in behavior. The memory on the new machine is quite a
good bit more than the old. Current memory usage right now as I speak is 22%
and it is using no swap memory whatsoever.
The machine uses a hardware RAID controller.
I tested to see if it could be an interaction between mysqld and NFS by turning
mysqld off and trying a copy. There was no change in behavior.
The file system on both the old and new system is ext3.
I'm not sure how to check the interrupt of the RAID controller. An ifconfig
gives the interrupt for eth0 as 22 and eth1 as 21.
I have unplugged eth1 to see if that is it and there is no change in behavior.
I increased the NFS threads from the default of 8 to 32 and restarted NFS.
However, based on nfsstat I don't think even needed to do that. It shows there
has been 0 retrans or authrefreshes.
=================
The problem at this point seems to be the NFS write. I've googled for this and
I come up with problems that have a regular read speed, but a write spee
measured in KB/s, which is MUCH slower than I have (perhaps I should be
thankful! :-) ). Most of what I have seen also tends to be for kernel <
2.6.12. I'm at 2.6.27.37.
I'm very much hoping that something here triggers something for someone.
Thanks, folks, for taking a look at things. I appreciate it!
---
Richard 'Doc' Kinne, [KQR]
American Association of Variable Star Observers
<rkinne @ aavso.org>
_______________________________________________
bblisa mailing list
[email protected]
http://www.bblisa.org/mailman/listinfo/bblisa
