konus - so I replaced the switch with a dhcp router and set it's IP to 192.168.0.50
Set the dhcp range to 192.168.0.1 to .40
Set the static IP for my laptop to 192.168.0.1
Started the tftp app on the laptop
Seeing the addresses of the laptop and KD20 on the dhcp list
Rather than shimmering, the power LED now starts solid, flashes, goes solid - then RED.
Settings:
How long should it take for the NAS to download the 2 files?
Also noteworthy - cannot ping NAS ...
The files are only a few MB so should be relatively quick, maybe a minute at most. Just to be clear I've not used / tried this method myself, it is just what the uboot sequence of commands show. If the loading of uImage and rdimg.gz works you should in theory end up with a system running some sort of minimal linux from RAM. Then there still is the issue of how to access this minimal system and re-flash stock. So the chance of this working is quite slim and things would be much easier once you have a working serial / uart connection.
If there is no output on the serial port then probably U-Boot cannot even be loaded and the only thing which could potentially work is using the SATA recovery method where the SoC directly loads U-Boot from a disk. TFTP recovery can only work if U-Boot on flash is still intact.
Yes sata boot would be needed to fix a broken uboot. However based on @TechGuy description the nas still beeps and the led changes colour which seems to indicate that uboot is still OK. So the priority should be to get the serial connection working by trying a different USB to UART converter and checking the wires.
Set the switch to 3.3V & hooked things up as follows:
Conv.....KD20
Gnd......Gnd
Tx.........Rx
Rx.........Tx
Booted Linux on the laptop and ran Minicom from a terminal window with 115200, N81 defaults and turned off hardware flow control (software flow control defaults off). Also set the serial device to /dev/ttyUSB0 . Emulation defaults to VT102.
With the HDD & fan removed, plugged in the KD20 and hit the power button. The power LED started solid blue, started to shimmer, the chirp sounded, then the power LED went red (that's consistent).
Nothing came through the serial connection (still).
Keyboard strokes produce a Tx LED response on the converter, but no Rx response.
This is not looking good ... like before ... but hope is dwindling
The USB, HDD & TFTP methods all failed as well. Is it something I'm missing?
What do you mean with that? You should connect only Gnd, Tx and Rx. Further you can test your converter, if you short Rx and Tx you should have a local echo.
Mijzelf, what I was trying to convey was that the Conv(erter) serial port pins below were connected to the corresponding KD20 serial port pins. Sorry about the confusion.
Shorted the Rx/Tx pins and the characters are looping back to the terminal.
When I reconnect to the KD20 - nothing.
Seems the NAS serial port is truly non-function now
OK, got my hands on a KD20 and successfully flashed the intermediate firmware. Powered down the NAS and came back to it tonight with the hopes of uploading the sysupgrade and ubinized files.
When I powered up the NAS it was looking good - the power LED was flickering blue, the HDD LEDs came on blue, the NAS chirped, the power LED went solid blue - and then a few seconds later, all the LEDs went out!?
The power LED came back on flickering blue - then it went solid - but the HDD LEDs are both out!?
The last time I had the NAS powered on all three LEDs were on and everything looked normal.
What happened!?
Where to from here ... ?
Ok, got OpenWrt successfully installed on the KD20, have the RAID1 built and have created a share using the LUCI GUI. Trying to access the share from Windows 11, it can't see the NAS, and from Linux I get:
Mijzelf, thank you for your reply. I'm in unfamiliar territory here. Looking at the end of the system log I'm seeing these related events but have no idea what is causing them, least of all why the ksmbd server doesn't run:
Wed Mar 8 23:55:12 2023 daemon.info ksmbd: triggering kill_server
Wed Mar 8 23:55:12 2023 kern.err kernel: [ 25.468193] ksmbd: kill_server_store:480: kill command received
Wed Mar 8 23:55:12 2023 daemon.notice ksmbd: Starting Ksmbd userspace service.
Wed Mar 8 23:55:12 2023 kern.err kernel: [ 25.714019] ksmbd: ksmbd_tcp_set_interfaces:596: Unknown interface: �duqRJOXF��1a
Wed Mar 8 23:55:12 2023 kern.err kernel: [ 25.721623] ksmbd: ksmbd_tcp_set_interfaces:596: Unknown interface:
...
Sun Mar 12 11:49:49 2023 daemon.info ksmbd: triggering kill_server
Sun Mar 12 11:49:49 2023 kern.err kernel: [ 105.522772] ksmbd: kill_server_store:480: kill command received
Sun Mar 12 11:49:49 2023 daemon.notice ksmbd: Starting Ksmbd userspace service.
jdwl1o1, thanks for your recommendation, but given the smaller size of ksmbd I'd like to put a bit of effort into seeing if I can get that working before I move on to samba ...
For clarity, the more things you can figure out how to turn off without breaking openwrt, the better the speeds for file transfers
Firewall is turned off by: system--> startup --> scroll down, firewall , stop, then disable --> then reboot
ipv6 is turned off by: network --> interfaces --> edit --> advanced settings -- > uncheck "Use builtin IPv6-management" --> save --> save & apply
bridge mode for lan: network --> interfaces --> edit --> physical settings --> uncheck "Bridge interfaces" , not sure what it was bridging to... --> save --> save & apply
removed the unknown interface that wasn't doing anything...because i don't know what that is.
also turned off Enable SYN-flood protection
network -->firewall --> general settings --> uncheck "enable SYN-flood protections"
if firewall was off, prob wasn't doing anything anyways
speeds closer to 15mb/s from 7-9mb/s . also smbd stopped crashing/closing...nvm still crashes with enough time..not sure what that is
Hi codetesstr, thanks for sharing your tips for improving throughput. Having applied these settings to my KD20 - I'm still seeing the device lock up requiring a hard reboot to clear the problem.
Any other thoughts, ideas, suggestions on how to keep things humming during large file transfers?
Correction - the NAS isn't hanging and does not require a reboot. Disconnecting the LAN cable and reconnecting it clears the problem ... the dead connectivity comes right back and the file copy continues!?
Have you tried replacing the LAN cable and ensuring it is at least a Cat 5 cable?
If the cable checks out and you have net-tools installed provide the output of the ifconfig command. You should pay particular attention to TX errors and RX errors.
Hi Ray, thanks for the suggestion. I've got a short Cat 5e cable installed between the NAS and the switch and a Cat 6 run to the workstation that's sending the files. The network is a 1000mbps. I've changed out the cable between the NAS and the switch to a longer 5e and will try some various combinations of interconnection between the stations and the router & the switch and will post the results ...
Update - tried various combinations - no change. Connectivity is still lost at some point during a long copy / backup. Is there any way of determining what's triggering the failure?