If you want to switch from passive to active benchmarking mode you could provide information from htop (are processes bottlenecked by CPU? Cores 0-3 are the little A7 ones and the beefy A15 are 4-7) and also 'iostat 5' while running such a transfer.
Hmm... iostat output and htop don't match (since iotop reports no %user activity but only %system and %iowait). Also it seems the master smbd process is bottlenecked by CPU utilizing a single A15 core (do you access SMB shares as root?)
Can you please provide the output from 'testparm' and 'armbianmonitor -u'?
This is a debian jessie image, apt-cache search armbianmonitor doesn't find a package.
Ok, then it's not worth the efforts. I'm only interested in performance of this specific OMV image (also Debian Jessie based but containing all Armbian and also new OMV optimizations for XU4): https://sourceforge.net/projec…oidxu4_4.9.28.7z/download
I won't start to waste time with older/unoptimized OS images (settings matter, kernel matters)
That image is only flashable to SD-Cards and my XU4 has only an eMMC media atm.
There seems to be a workaround available in the meantime. But requires overwriting the eMMC with another image (since the u-boot Hardkernel put on the eMMC by default is horribly outdated and doesn't even support ext4): https://forum.armbian.com/inde…mmc/&page=3#comment-33345
Thanks for the link. I'll see if I can make sense of it 
Ok, I've found an older 16 GB micro SD-Card lying around... I've flashed the
https://sourceforge.net/projec…oidxu4_4.9.28.7z/download
image and reconfigured everything
htop
iostat 5
Device: tps kB_read/s kB_wrtn/s kB_read kB_wrtn
mmcblk1 0.00 0.00 0.00 0 0
sda 18.80 0.80 18848.00 4 94240
dm-0 22.20 0.80 20688.00 4 103440
avg-cpu: %user %nice %system %iowait %steal %idle
2.09 0.00 8.58 17.90 0.00 71.43testparm
root@odroid-xu4:~# testparm
Load smb config files from /etc/samba/smb.conf
WARNING: The "null passwords" option is deprecated
Processing section "[storage]"
Processing section "[crypt]"
Loaded services file OK.
Server role: ROLE_STANDALONE
Press enter to see a dump of your service definitions
# Global parameters
[global]
workgroup = HIGHEND
server string = %h server
pam password change = Yes
passwd program = /usr/bin/passwd %u
passwd chat = *Enter\snew\s*\spassword:* %n\n *Retype\snew\s*\spassword:* %n\n *password\supdated\ssuccessfully* .
syslog = 0
syslog only = Yes
log file = /var/log/samba/log.%m
max log size = 1000
min receivefile size = 16384
socket options = TCP_NODELAY IPTOS_LOWDELAY
load printers = No
printcap name = /dev/null
disable spoolss = Yes
dns proxy = No
panic action = /usr/share/samba/panic-action %d
idmap config * : backend = tdb
create mask = 0777
directory mask = 0777
aio read size = 16384
aio write size = 16384
use sendfile = Yes
write cache size = 524288
printing = bsd
[storage]
path = /srv/dev-disk-by-label-STORAGE/storage
valid users = highend
write list = highend
read only = No
create mask = 0664
force create mode = 0664
directory mask = 0775
force directory mode = 0775
inherit acls = Yes
hide special files = Yes
[crypt]
path = /srv/dev-disk-by-label-CRYPT/crypt
valid users = highend
write list = highend
read only = No
create mask = 0664
force create mode = 0664
directory mask = 0775
force directory mode = 0775
inherit acls = Yes
hide special files = Yesarmbianmonitor -u
http://sprunge.us/AVZE
A picture how the transfer to the crypted volume looks like
If you need any more info...
Thanks for the link. I'll see if I can make sense of it
Ok, I've found an older 16 GB micro SD-Card lying around... I've flashed the
https://sourceforge.net/projec…oidxu4_4.9.28.7z/download
image and reconfigured everything
htop
iostat 5
Device: tps kB_read/s kB_wrtn/s kB_read kB_wrtn
mmcblk1 0.00 0.00 0.00 0 0
sda 18.80 0.80 18848.00 4 94240
dm-0 22.20 0.80 20688.00 4 103440
avg-cpu: %user %nice %system %iowait %steal %idle
2.09 0.00 8.58 17.90 0.00 71.43testparm
root@odroid-xu4:~# testparm
Load smb config files from /etc/samba/smb.conf
WARNING: The "null passwords" option is deprecated
Processing section "[storage]"
Processing section "[crypt]"
Loaded services file OK.
Server role: ROLE_STANDALONE
Press enter to see a dump of your service definitions
# Global parameters
[global]
workgroup = HIGHEND
server string = %h server
pam password change = Yes
passwd program = /usr/bin/passwd %u
passwd chat = *Enter\snew\s*\spassword:* %n\n *Retype\snew\s*\spassword:* %n\n *password\supdated\ssuccessfully* .
syslog = 0
syslog only = Yes
log file = /var/log/samba/log.%m
max log size = 1000
min receivefile size = 16384
socket options = TCP_NODELAY IPTOS_LOWDELAY
load printers = No
printcap name = /dev/null
disable spoolss = Yes
dns proxy = No
panic action = /usr/share/samba/panic-action %d
idmap config * : backend = tdb
create mask = 0777
directory mask = 0777
aio read size = 16384
aio write size = 16384
use sendfile = Yes
write cache size = 524288
printing = bsd
[storage]
path = /srv/dev-disk-by-label-STORAGE/storage
valid users = highend
write list = highend
read only = No
create mask = 0664
force create mode = 0664
directory mask = 0775
force directory mode = 0775
inherit acls = Yes
hide special files = Yes
[crypt]
path = /srv/dev-disk-by-label-CRYPT/crypt
valid users = highend
write list = highend
read only = No
create mask = 0664
force create mode = 0664
directory mask = 0775
force directory mode = 0775
inherit acls = Yes
hide special files = Yesarmbianmonitor -u
http://sprunge.us/AVZE
A picture how the transfer to the crypted volume looks like
If you need any more info...
Thank you for the data (still htop and iostat not really consistent). Next step is to eliminate Samba with a local iozone bench.
Please cd to your LUKS directory and then use first iozone call from https://forum.armbian.com/inde…findComment&comment=15265 (iozone -e -I -a -s 100M -r 4k -r 16k -r 512k -r 1024k -r 16384k -i 0 -i 1 -i 2).
And run in parallel in one shell 'htop' (and have an eye there whether specific processes max out one CPU core) and run in another shell 'iostat 5' and stop it with [ctrl-c] only after iozone has finished. And then results from both calles via pastebin.com would be great.
No cpu core is maxed out during the test. Highest load for one core was less than 50%
Samba itself is fine. On the unencrypted part of the external usb3 drive I get a constant
write rate for one large file at 110 MB/sec...
iostat pastebin
https://pastebin.com/g8k2uAiG
iozone output
https://pastebin.com/HT9LCKrT
Picture of htop
Ok, low performance is unrelated to Samba (and settings) and to OMV too. With iozone you also get just ~18MB/s and iostat reports being somewhat busy in %system and twice as much in %iowait.
The only advise I could give is to try out performance governor and see whether performance improves (cpufreq-set -g performance). If not you should ask over in Hardkernel forum referencing this thread wrt LUKS implementation (to me it seems there's something single threaded but the tasks jumps between CPU cores).
BTW: I used your armbianmonitor output to illustrate the difference between a smartphone SoC and a NAS SoC: Helios4 -- community developed ARM 4 bay NAS device received important upgrade
Sie haben noch kein Benutzerkonto auf unserer Seite? Registrieren Sie sich kostenlos und nehmen Sie an unserer Community teil!