I have like 500GB copying over to my NAS to disk2 (and the parity drive) which is writing around the normal 20MB/s (since I don't have a cache drive).

 

When I try to get a directory listing off of disk1 it takes forever.  IT takes forever to play an MP3 from iTunes from disk1 as well. 

 

This doesn't make sense to me because disk1 is on a different channel than disk2 and the parity drive.  They all 3 are using the sata ports on the motherboard.

 

So is this a problem with my motherboard?  Do I need to upgrade it to a server grade motherboard?  Looking at the diagram I see there is a PCIE bus all the pci-e devices share.. perhaps this is a weaker bus on this motherboard than others?  (I'm using ASUS P5GC-MX/1333.)

 

I doubt this could be a software or cpu problem because the bottleneck is the pci-e bus and hard drives.. the cpu is plenty fast.. much more than it needs to be .. core2duo 2.6ghz.

 

I'm running 5.0 rc3 and Snow Leopard 10.6.8.  The 500gb file copy is via AFP and the small reads of itunes music are also via AFP.

 

Post a syslog please.

 

There's no reason that this should be a bandwidth issue. UnRAID doesn't even saturate a SATA I connection, and your motherboard supports SATA II.

 

It seems more likely that it's an issue with your network.

Post a syslog please.

 

 

May 29 06:07:26 Galaxy emhttp: Start array...

May 29 06:07:26 Galaxy kernel: mdcmd (16): start STOPPED

May 29 06:07:26 Galaxy kernel: unraid: allocating 18360K for 1280 stripes (3 disks)

May 29 06:07:26 Galaxy kernel: md1: running, size: 1953514552 blocks

May 29 06:07:26 Galaxy kernel: md2: running, size: 1953513492 blocks

May 29 06:07:26 Galaxy emhttp: shcmd (181): udevadm settle

May 29 06:07:26 Galaxy emhttp: shcmd (182): /usr/local/sbin/emhttp_event array_started

May 29 06:07:26 Galaxy emhttp_event: array_started

May 29 06:07:26 Galaxy emhttp: Mounting disks...

May 29 06:07:26 Galaxy emhttp: shcmd (183): mkdir /mnt/disk2

May 29 06:07:26 Galaxy emhttp: shcmd (184): mkdir /mnt/disk1

May 29 06:07:26 Galaxy emhttp: shcmd (185): set -o pipefail ; mount -t reiserfs -o user_xattr,acl,noatime,nodiratime /dev/md2 /mnt/disk2 |& logger

May 29 06:07:26 Galaxy emhttp: shcmd (186): set -o pipefail ; mount -t reiserfs -o user_xattr,acl,noatime,nodiratime /dev/md1 /mnt/disk1 |& logger

May 29 06:07:26 Galaxy kernel: mdcmd (17): check CORRECT

May 29 06:07:26 Galaxy kernel: md: recovery thread woken up ...

May 29 06:07:26 Galaxy kernel: REISERFS (device md2): found reiserfs format "3.6" with standard journal

May 29 06:07:26 Galaxy kernel: REISERFS (device md2): using ordered data mode

May 29 06:07:26 Galaxy kernel: REISERFS (device md2): journal params: device md2, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

May 29 06:07:26 Galaxy kernel: REISERFS (device md2): checking transaction log (md2)

May 29 06:07:26 Galaxy kernel: REISERFS (device md1): found reiserfs format "3.6" with standard journal

May 29 06:07:26 Galaxy kernel: REISERFS (device md1): using ordered data mode

May 29 06:07:26 Galaxy kernel: md: recovery thread has nothing to resync

May 29 06:07:26 Galaxy kernel: REISERFS (device md1): journal params: device md1, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30

May 29 06:07:26 Galaxy kernel: REISERFS (device md1): checking transaction log (md1)

May 29 06:07:26 Galaxy kernel: REISERFS (device md2): Using r5 hash to sort names

May 29 06:07:26 Galaxy kernel: REISERFS (device md1): Using r5 hash to sort names

May 29 06:07:27 Galaxy emhttp: shcmd (187): chmod 770 '/mnt/disk2'

May 29 06:07:27 Galaxy emhttp: shcmd (188): chmod 770 '/mnt/disk1'

May 29 06:07:27 Galaxy emhttp: shcmd (189): chown nobody:users '/mnt/disk2'

May 29 06:07:27 Galaxy emhttp: shcmd (190): chown nobody:users '/mnt/disk1'

May 29 06:07:27 Galaxy emhttp: shcmd (191): mkdir /mnt/user

May 29 06:07:27 Galaxy emhttp: shcmd (192): /usr/local/sbin/shfs /mnt/user -disks 6 -o noatime,big_writes,allow_other,default_permissions,use_ino

May 29 06:07:27 Galaxy emhttp: shcmd (193): crontab -c /etc/cron.d -d &> /dev/null

May 29 06:07:27 Galaxy emhttp: shcmd (194): /usr/local/sbin/emhttp_event disks_mounted

May 29 06:07:27 Galaxy emhttp_event: disks_mounted

May 29 06:07:27 Galaxy emhttp: shcmd (195): :>/etc/samba/smb-shares.conf

May 29 06:07:27 Galaxy emhttp: shcmd (196): cp /etc/netatalk/AppleVolumes.default- /etc/netatalk/AppleVolumes.default

May 29 06:07:27 Galaxy emhttp: Start SMB...

May 29 06:07:27 Galaxy emhttp: shcmd (197): /etc/rc.d/rc.samba start |& logger

May 29 06:07:27 Galaxy logger: Starting Samba:  /usr/sbin/nmbd -D

May 29 06:07:27 Galaxy logger:                  /usr/sbin/smbd -D

May 29 06:07:27 Galaxy emhttp: shcmd (198): ps axc | grep -q rpc.mountd

May 29 06:07:27 Galaxy emhttp: _shcmd: shcmd (198): exit status: 1

May 29 06:07:27 Galaxy emhttp: Start AFP...

May 29 06:07:27 Galaxy emhttp: shcmd (199): /etc/rc.d/rc.atalk start |& logger

May 29 06:07:27 Galaxy logger: starting appletalk daemons:  cnid_metad afpd

May 29 06:07:27 Galaxy emhttp: shcmd (200): cp /etc/avahi/services/afp.service- /etc/avahi/services/afp.service

May 29 06:07:27 Galaxy emhttp: shcmd (201): cp /etc/avahi/services/smb.service- /etc/avahi/services/smb.service

May 29 06:07:27 Galaxy emhttp: shcmd (202): /usr/local/sbin/emhttp_event svcs_restarted

May 29 06:07:27 Galaxy emhttp_event: svcs_restarted

May 29 06:07:27 Galaxy avahi-daemon[1930]: Files changed, reloading.

May 29 06:07:27 Galaxy avahi-daemon[1930]: Loading service file /services/afp.service.

May 29 06:07:27 Galaxy avahi-daemon[1930]: Loading service file /services/smb.service.

May 29 06:07:28 Galaxy avahi-daemon[1930]: Service "Galaxy-SMB" (/services/smb.service) successfully established.

May 29 06:07:28 Galaxy avahi-daemon[1930]: Service "Galaxy" (/services/afp.service) successfully established.

May 29 06:14:08 Galaxy in.telnetd[2205]: connect from 192.168.1.115 (192.168.1.115)

May 29 06:14:10 Galaxy login[2206]: invalid password for 'geekess'  on '/dev/pts/0' from 'JennMac'

May 29 06:14:14 Galaxy login[2206]: ROOT LOGIN  on '/dev/pts/0' from 'JennMac'

May 29 06:21:07 Galaxy afpd[2163]: deletecurdir: error deleting .AppleDouble in " "

 

 

I see no errors in the syslog since I started the copy.  I think I started the copy of the 500gb around 7am.

 

Btw, I can go back further in the log if you need any information I didn't paste above.

Okay, I did a test. 

 

I have another computer on the network besides my Snow Leopard 10.6.8 client:  A Windows 7 machine, which accesses the unRaid server via SMB.

 

The 500GB copy is still proceeding from my Snow Leopard machine to the NAS via AFP (to disk2), and the reads from disk1 are still painfully slow on this machine.

 

However, the reads from disk1 from the Windows 7 machine (via SMB) are very fast (while this copy is going on).

 

So this tells me it's a problem with either AFP on the unRaid server or AFP on my Snow Leopard.

 

They are the same machines hardware-wise: same nic, memory, cpu, motherboard etc..

 

I am intending on upgrading to Lion soon and I heard that solves the slow write problems with SMB (that snow leopard has).  Hopefully I can replace AFP with SMB on lion and get the same read performance I am from the Windows 7 Machine while large copies are going on to other hard disks.

 

Anyways, its good to know it's not my hardware in the unRaid Server.