You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
After a little research found your driver, it helped influence my buying choices - new DS1819+ to replace ageing DX1511+ and DX510 combo, 2x QNAP QNA-UC5G1T for direct link workstation-server.
Right off the bat, THANK YOU. After 3 days of benchmarking file systems and network protocols machine is now production, I'm super happy with it, your driver is a crucial part of that.
TL;DR? Working great, maxed it out on large single file test against 6-drive RAID6. A few tiny bugs during install. SMB kills NFS across all tests Win10 - DS.
DS1819+ is Denverton Celeron. Comes with a PCIE slot can be either 10Gb card or M2 cache. Since i'm only interested in improving point to point transfers between workstation and the DS, not providing >2G bond to rest of network, wanted to keep options open for expansion card, so opted to try pair of QNAP's.
Installation went fine, i'm n00b at the prompt so went through manual install in package centre. After reboot new interface appeared OK under CP/Network. At first, Manual MTA showed only 1500 option in dropdown.
Installed other adapter on USB3 port on TB3 dock for Lenovo X1, direct Cat6 cable to DS.
Forced link speed to 5G on workstation end, seems driver is reporting incorrect capabilities to DS system - possible link speeds, maybe range allowed MTU also?:
Output of ethtool eth4 command
Settings for eth4:
Supported ports: [ TP MII ]
Supported link modes: 100baseT/Full
1000baseT/Full
Supported pause frame use: No
Supports auto-negotiation: Yes
Advertised link modes: 100baseT/Full
1000baseT/Full
Advertised pause frame use: No
Advertised auto-negotiation: Yes
Speed: 5000Mb/s
Duplex: Full
Port: MII
PHYAD: 0
Transceiver: internal
Auto-negotiation: on
Cannot get wake-on-lan settings: Operation not permitted
Current message level: 0x00000007 (7)
drv probe link
Link detected: yes
Forced MTU 9000 with ip link set eth4 mtu 9000 by ssh at NAS.
After reboot (I'm sure could be done less brutally at shell but again, I'm terminal n00b) the Manual MTU option under Network-Network Interface-Edit shows fully populated list 1500-9000 with everything in between.
Although it's on USB, I'm pretty sure you can't hotplug the adapter, at least for me it misbehaved and I needed a reboot of the DS to get stable connection back. Might well be doable somehow from the shell but beyond my scope, someone more expert with synology can maybe clear that up.
With a stable connection, I set about benchmarking SMB and NFS for transfers to and from Win10 and DS. Note, I had upgraded to 32GB onboard memory - in prior tests I observed up to 30.5GB memory used for cache so my tests were:
very large single file (1x64gb zip, 1x28Gb zip) - used to definitely exceed / be just within max cache
folder with 28GB of audio files each 3-12MB (20% of my data)
folder with 28GB of video files each 2-7GB (50% of my data)
Repeatedly tested transfer back and forth of these, first over SMB then over NFS.
Results:
Large single file first copy approx 300-350MB/s range, mostly 320MB/s sustained
Second copy of 28GB file (fully cached by DS) 430MB/s sustained (Yes!)
Audio files first copy approx 80-170MB/s range
Cached copy approx 200-280MB/s range
Video files first approx 175-250MB/s range
Cached copy approx 200-300MB/S range
Across all file sizes, NFS only every managed 100-150MB/s range, cached or uncached.
CPU utilisation under NFS 17-24%
CPU utilisation under SMB 40-45%
These are just first benchmarks, happy to run something specific if anyone wants, and have screenshots during all the above tests also if requested... learning more shell every day but please be patient if you need me to run CLI commands, be explicit :)
I'll post a follow-on in a few weeks, by then should have some feedback on stability and real world usage on vid production and machine backups.
After a little research found your driver, it helped influence my buying choices - new DS1819+ to replace ageing DX1511+ and DX510 combo, 2x QNAP QNA-UC5G1T for direct link workstation-server.
Right off the bat, THANK YOU. After 3 days of benchmarking file systems and network protocols machine is now production, I'm super happy with it, your driver is a crucial part of that.
TL;DR? Working great, maxed it out on large single file test against 6-drive RAID6. A few tiny bugs during install. SMB kills NFS across all tests Win10 - DS.
DS1819+ is Denverton Celeron. Comes with a PCIE slot can be either 10Gb card or M2 cache. Since i'm only interested in improving point to point transfers between workstation and the DS, not providing >2G bond to rest of network, wanted to keep options open for expansion card, so opted to try pair of QNAP's.
Installation went fine, i'm n00b at the prompt so went through manual install in package centre. After reboot new interface appeared OK under CP/Network. At first, Manual MTA showed only 1500 option in dropdown.
Installed other adapter on USB3 port on TB3 dock for Lenovo X1, direct Cat6 cable to DS.
Forced link speed to 5G on workstation end, seems driver is reporting incorrect capabilities to DS system - possible link speeds, maybe range allowed MTU also?:
Output of
ethtool eth4
commandForced MTU 9000 with
ip link set eth4 mtu 9000
by ssh at NAS.After reboot (I'm sure could be done less brutally at shell but again, I'm terminal n00b) the Manual MTU option under Network-Network Interface-Edit shows fully populated list 1500-9000 with everything in between.
Although it's on USB, I'm pretty sure you can't hotplug the adapter, at least for me it misbehaved and I needed a reboot of the DS to get stable connection back. Might well be doable somehow from the shell but beyond my scope, someone more expert with synology can maybe clear that up.
With a stable connection, I set about benchmarking SMB and NFS for transfers to and from Win10 and DS. Note, I had upgraded to 32GB onboard memory - in prior tests I observed up to 30.5GB memory used for cache so my tests were:
Repeatedly tested transfer back and forth of these, first over SMB then over NFS.
Results:
Large single file first copy approx 300-350MB/s range, mostly 320MB/s sustained
Second copy of 28GB file (fully cached by DS) 430MB/s sustained (Yes!)
Audio files first copy approx 80-170MB/s range
Cached copy approx 200-280MB/s range
Video files first approx 175-250MB/s range
Cached copy approx 200-300MB/S range
Across all file sizes, NFS only every managed 100-150MB/s range, cached or uncached.
CPU utilisation under NFS 17-24%
CPU utilisation under SMB 40-45%
These are just first benchmarks, happy to run something specific if anyone wants, and have screenshots during all the above tests also if requested... learning more shell every day but please be patient if you need me to run CLI commands, be explicit :)
I'll post a follow-on in a few weeks, by then should have some feedback on stability and real world usage on vid production and machine backups.
Thanks again @bb-qq, domo arigato gozaimasu
Description of your products
Description of your environment
Output of
ifconfig -a
commandThe text was updated successfully, but these errors were encountered: