-
Notifications
You must be signed in to change notification settings - Fork 189
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
DS412+ (Atom D2700) #7
Comments
Since the controller chip of B07VNFLTLD is RTL8156, the dongle should work with my driver. If you prefer SFP+, QSW-308-1C is another choice for this dongle. |
Thank you for that information! I will give it a shot and report back.
…On Thu, Jan 2, 2020, 22:24 bb-qq ***@***.***> wrote:
Since the controller chip of B07VNFLTLD is RTL8156, the dongle should work
with my driver.
However please note that NBaseT(2.5G/5G) needs dedicated transceivers for
it. Probably 10G-SFP-T does not support 2.5/5G.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#7?email_source=notifications&email_token=AFJ35YH6F4ZDNOXUSQMGA23Q32VYFA5CNFSM4KB6UB22YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEIAFQQY#issuecomment-570447939>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFJ35YFHETBIGVQ5TST4G53Q32VYFANCNFSM4KB6UB2Q>
.
|
Description of the problemConfused about how to configure the IP. Should it be listed as a DSM Network Interface? r8152-cedarview-2.12.0-1.spk is running, B07VNFLTLD is recognized per lsusb, but I have no idea where/how to access the Web UI for step 3. Description of your productsDS412+ Description of your environmentPlanned: Mikrotik RS305-1G-4S+IN switch with an ipolex 10G-SFP-T transceiver. ### Output of lsusb command |__usb1 1d6b:0002:0310 09 2.00 480MBit/s 0mA 1IF (ehci_hcd 0000:00:1 a.7) hub |
Did you run |
I did run the command, but it did not list eth2 after "ifconfig -a".
…On Fri, Jan 17, 2020, 04:41 bb-qq ***@***.***> wrote:
Did you run ifconfig eth2 up and is it succeeded?
Please confirm the result of ifconfig -a.
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#7?email_source=notifications&email_token=AFJ35YCWD5IXO5PVDUNBJVDQ6F4NLA5CNFSM4KB6UB22YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJHDODI#issuecomment-575551245>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFJ35YEWLNFPWGII3TCP2QTQ6F4NLANCNFSM4KB6UB2Q>
.
|
Hmm, it's weird. VID/PID shown in the result of So could you let me know the result of That can be obtained by the following steps:
|
[77414.866474] usb 8-2: new SuperSpeed USB device number 3 using
etxhci_hcd-170202
[77414.894713] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2 ep
2 with no TDs queued?
[77414.903531] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2 ep
3 with no TDs queued?
[77414.912282] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2 ep
6 with no TDs queued?
[77416.023156] usb 8-2: reset SuperSpeed USB device number 3 using
etxhci_hcd-170202
[77416.580569] r8152 8-2:1.0 eth2: v2.12.0 (2019/04/29)
[77416.585547] r8152 8-2:1.0 eth2: This product is covered by one or more
of the following patents:
US6,570,884, US6,115,776, and US6,327,625.
…On Fri, Jan 17, 2020 at 11:04 AM bb-qq ***@***.***> wrote:
Hmm, it's weird. VID/PID shown in the result of lsusb is certainly
supported by the driver.
So could you let me know the result of dmesg? The required part is only
about detection of the dongle.
That can be obtained by the following steps:
1. detach the ethernet dongle
2. run dmesg
3. attach the ethernet dongle
4. run dmesg and check increased part from previous output
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#7?email_source=notifications&email_token=AFJ35YEGCMPR3E7RUB7GQ6DQ6HJJ7A5CNFSM4KB6UB22YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJIESBY#issuecomment-575686919>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFJ35YGLAPDA7P4WZITETLTQ6HJJ7ANCNFSM4KB6UB2Q>
.
|
Also, I just ran "ifconfig eth2 up" again and received this message:
"SIOCSIFFLAGS: Operation not permitted". I then elevated to root and ran it
again. As root, no error message was generated. Logging out as root and
repeating the command as admin once again led to the SIOCSIFFLAGS message.
I originally ran the command as root, so perhaps this has happened both
times. Risky I guess, but I *just* archived my entire NAS anyway.
…On Fri, Jan 17, 2020 at 8:55 PM Seth West ***@***.***> wrote:
[77414.866474] usb 8-2: new SuperSpeed USB device number 3 using
etxhci_hcd-170202
[77414.894713] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2
ep 2 with no TDs queued?
[77414.903531] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2
ep 3 with no TDs queued?
[77414.912282] etxhci_hcd-170202 0000:04:00.0: WARN Event TRB for slot 2
ep 6 with no TDs queued?
[77416.023156] usb 8-2: reset SuperSpeed USB device number 3 using
etxhci_hcd-170202
[77416.580569] r8152 8-2:1.0 eth2: v2.12.0 (2019/04/29)
[77416.585547] r8152 8-2:1.0 eth2: This product is covered by one or more
of the following patents:
US6,570,884, US6,115,776, and US6,327,625.
On Fri, Jan 17, 2020 at 11:04 AM bb-qq ***@***.***> wrote:
> Hmm, it's weird. VID/PID shown in the result of lsusb is certainly
> supported by the driver.
>
> So could you let me know the result of dmesg? The required part is only
> about detection of the dongle.
>
> That can be obtained by the following steps:
>
> 1. detach the ethernet dongle
> 2. run dmesg
> 3. attach the ethernet dongle
> 4. run dmesg and check increased part from previous output
>
> —
> You are receiving this because you authored the thread.
> Reply to this email directly, view it on GitHub
> <#7?email_source=notifications&email_token=AFJ35YEGCMPR3E7RUB7GQ6DQ6HJJ7A5CNFSM4KB6UB22YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJIESBY#issuecomment-575686919>,
> or unsubscribe
> <https://github.com/notifications/unsubscribe-auth/AFJ35YGLAPDA7P4WZITETLTQ6HJJ7ANCNFSM4KB6UB2Q>
> .
>
|
Can you assign an IP address after that? |
YES! I don't know what changed, but it now shows up as LAN3 and I can
assign the IP manually. Thank you so much for your work and for helping me
get this to work for my system. I've connected it to the Mikrotik switch
via IPOLEX 10G-SFP-T to CAT6 and DSM reports it as "2500 Mbps full-duplex,
MTU 1500". I'm still waiting on an SFP+ NIC for my computer, tho.
This is a huge accomplishment on your part. I hope people recognize the
work you've done. Do you have a donation page anywhere?
…On Sun, Jan 19, 2020 at 6:34 AM bb-qq ***@***.***> wrote:
Can you assign an IP address after that?
—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
<#7?email_source=notifications&email_token=AFJ35YGFAD2YLK6F6TOSFK3Q6Q3CPA5CNFSM4KB6UB22YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEJKP5SQ#issuecomment-575995594>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/AFJ35YG64YW6MZLMPYRXC6LQ6Q3CPANCNFSM4KB6UB2Q>
.
|
It's my pleasure.
I have created a donation page for the experiment. Please take a look if you feel like it. |
Excellent, thank you. I have one more issue that I resolved myself. This is more of an FYI with a question at the end. Here's how I left my network yesterday. For brevity, all transceivers (Tx) are IOPLEX 10G-SFP-T models.
I had just connected the system together and configured the IP for LAN3 in DSM, verified that the IP was reporting, and then tried to get iperf3 running. I failed to get iperf3 to do anything except hang in terminal, and then left the system for ~24hours. Upon return, LAN3 had disappeared. The lights on the RTL8156 device were still flashing (white and orange, indicating 2.5GbE connection?). My system did not experience a power failure. It's connected to a UPS and neither system reported an outage. I was able to restore the device to active state by doing the following steps, though I do not know if all steps were required. 1) I stopped and started the RTL driver through Package Center. 2) I ran Could we perhaps have a user-script in Task Scheduler that checks that eth2 is active and, if not, run the command? EDIT: Some output. ~$ dmesg | grep eth |
You may want to try the standard Gigabit Ethernet connection first. If it works, there seems to be a problem with the NBASE-T environment or thermal issue. If it doesn't work, try removing other USB devices. If that doesn't work too, reloading the driver may be the last solution: Also, thank you for your donation! |
After my last post, I started using the RTL as the only gateway. I changed it's IP to my default IP for the NAS and unplugged LAN1. The connection has been reliable for the past couple of days. It sustained itself even through a scheduled sleep cycle. Thank you for the script! I don't believe the rest of this has anything to do with your work. I'm really just sharing it because it confirms your suspicion, and explains why I haven't verified that the setup hit the intended performance goal. I finally had time to get iperf3 working and it's confirming your suspicion about the Tx. After you brought up your concern with 10GBase-T transceivers operating as 2.5/5 GBase-T speeds, I was sure to purchase a model of Tx that supposedly supports 2.5/5/10 GBase-T. Considering I'm not even hitting 100 Mbps speeds when it's connected to anything other than a matching Tx, I think it's fairly conclusive that it's "support" means that there is a connection...but only just barely. Here's the summary. When I say "no Tx", that means I'm using the built-in RJ45 console port: Further testing the Tx, I connected the PC (Tx) to the RTL directly. Windows reported 10 Gbe...so that's not to be trusted... and iperf resulted in 73 to 400 Mbps (avg. of 264). Guess I'll look for a 2.5GBase-T Tx and hope for the best. It's proving a difficult search, so I'd appreciate any recommendations if you have them. |
Hmm, it is a difficult situation... What is your MTU value? Because of USB based ethernet drivers consume more CPU and the USB protocol has a large amount of overhead, enabling Jumbo-frame is required to achieve wire-speed in some cases.
Can you fix link-speed to 1G form the Device Manager on the PC side? |
I connected the PC (Tx) to RTL again for these values: I haven't found a way to change the link-speed of the Mellanox card. I varied the jumbo packet size with the same settings on each side:
I then varied the Mellanox Rx/Tx Buffers (256 to 4096 range per). Starting point: Rx/Tx buffer 4096/2048
Then to mix things up...
Returned all to 1500 MTU, buffers to default 4096/2048 |
I found a specification sheet for the Tx I've been using. Despite what I had previously read, it is not compatible with 2.5 / 5 GBase-T. I've ordered a FS(.com) 10G-T-SFP-PLUS that explicitly states compatibility with 1/2.5/5/10 GBase-T in the specification sheet. |
Do you have another PC? Confirming PC to PC communication should help to isolate the problem. Also, I noticed that Connectx-2 does not support NBASE-T because the device was released before publishing IEEE 802.3bz. We can only use 1Gigabit link under direct connection. |
yeah~ after "ifconfig eth3 up", goto your dsm webpage and go control panel -> network panel. |
I received the new Tx (10G-T-SFP-PLUS) today. I installed it into the Mikrotik, then connected it to the RTL (Syno) and ran iperf. No dice. ~88 Mbps to the Mellanox/desktop.
I have a laptop with GbE and USB 3.0 ports, but nothing I can plug the Mellanox card into. I connected the laptops GbE port into the RTL (Syno) and received GbE performance as confirmed by NIC status and iperf3 to my desktop. I then moved the RTL to a USB 3 port on my laptop and attached it to the new Tx in the Mikrotik. Iperf3 from my desktop... 69 Mbps. Sigh.
That is a good catch. I haven't established a connection with the CRS305 interface yet, but I hope to find a way to run speed tests directly to/from it when I have time. |
This result seems to indicate that CRS305 has an issue with 2.5G/5G, but this user reported it works. It's strange... By the way, are you using the latest firmware? If not, it might be worth to update. |
Well, I've now acquired the S+RJ10 transceiver and given it a whirl. It's at least negotiating and saturating 1GbE speeds now, so that's an improvement. I've also upgraded the firmware on the MikroTik CRS305. Per MikroTik, the CRS305 with S+RJ10 should auto-negotiate 1/2.5/5/10 GBase-T. The cable works at 10G speeds - verified. Not sure what's going on now... If I turn auto-negotiate off on the CRS305 then I lose the options for 2.5/5 G when using the WinBox gui. Not sure if I can get it run via command line but I'll try that next time I'm digging into it. BTW - it seems there's some recurrence of the driver package not responding after the attached RTL device has been left disconnected for a time. The method to restart the driver from before still works fine (restart driver, issue the 'ifconfig eth2 up" command as root). |
I can confirm using S+RJ10 Mikrotik transceiver I get 2.5g auto-negotiated. please notethat unless you have the latest version of this transceiver you can not use jumbo frames at all, its a known bug. |
Thank you for that information. I will turn them off when I return to debugging. |
Did it go well? Also please make sure using the front port because some users reported instability issues when they use rear ports. |
Hey! Thanks for checking in. I have not yet returned to this project - even with the quarantine...! Your comment about the front USB port is not feasible. That is a USB 2.0 port. As such, it is limited to a theoretical maximum of 480 Mbps; a far cry from the target 2500 Mbps. |
I concluded Synology devices have a signal quality issue on rear USB ports from the fact while the front USB port is soldered on the main PCB, the rear USB port is not. (DS918+ tear down) I will update the document to mention this. |
ContextModel(platform): DS412+ (Cedarview) with 2048 MB RAM upgrade
Advice: Don't spend your money on this if you have a older DS412+ model or don't get your hopes up lsusb output
iperf3 testsConnected to a Mikrotik S+RJ10 v1 in a Miktrotik CRS309.
Nothing to write home about. |
Poked at this again. It's remarkable that this little NAS is still kicking after 11 years of continuous operation. Revisiting #7 Constants:
Design of Experiments:
Learnings / Tips
Results Issues
|
Description of the problem
I am not adept in the wizardly ways of creating drivers, and would very much like to have a 2.5 or 5Gbps network connection with my NAS.
Description of your products
DS412+
Linux 3.10.105 #24922 SMP Wed Jul 3 16:35:48 CST 2019 x86_64 GNU/Linux synology_cedarview_412+
DSM 6.2.2
Planned: CableCreation PN B07VNFLTLD.
Description of your environment
Planned: Mikrotik RS305-1G-4S+IN switch with an ipolex 10G-SFP-T transceiver.
Planned: Mellanox Connectx-2
CAT6
Output of
dmesg
commandOutput of
lsusb
commandThe text was updated successfully, but these errors were encountered: