v2.0.9rc1 Bug Reports and Comments - WS3 Firmware

[Stephen]

Latest WS3 Release Candidate: v2.0.9rc1

Click HERE to download WS3 firmware v2.0.9rc1 - Released 11/15/2023

FIXED/CHANGED
- Fixed undefined behavior caused by dash characters in switch name - rc1

NOTE:
This fix directly applies to changing the switch name to something like "switch-test-1", doing this caused severe delay's applying config's, switch lock-up, and unexpected behavior after reboots.
Due to the nature of the error, components of what was needed to fix this specific issue may have been responsible for other types of config changes causing similar error's. We will need feedback to verify if this is the case or not. During testing, we are now unable to replicate the behavior's described.
With any luck, this might solve the majority of stability issue's we're seeing on the WS3.

Some people have reported needing to default switch after upgrade to RC1. Probably more needed if you had a dash in the switch name.


Please let us know in this thread what you find when trying this firmware.

ENHANCEMENTS

KNOWN ISSUES

- STP/RSTP/MSTP Broken, disable it.

- If you have an OCP on a port try using a different port see if bad port or problem follows the cable.
- Also another issue just discovered inserting some if not all 1G SFP modules causes a reboot
Inserting 10G SFP+ modules work fine.
Hoping to have a new release in a couple weeks.

- WEB UI issues when not at 100% Zoom on browser especially on VLAN TAB
- Some language templates need help

Released 11/15/2023

[backtick]

I can confirm everything that did not work now does work! The GUI seems snappier too, great stuff

Only problem that does not seem to go away is the OCP. 1,6,7 48VH ports allocated always need a Reset, whether or not there's a cable connected. Once its reset there is never a problem and cable diagnostics are clean. Port 1 has never powered a radio and the others see max 40W so I am stumped to as why its happening after every reboot. Any advice? Thanks

ps. the switch is on the bench; the entries are chronological

Nov 16 03:41:11 switch[45]: Disabling PoE power on port 6 (48VH-AF5XHD-) due to Over Current Protection
Nov 16 03:41:11 switch[45]: Disabling PoE power on port 7 (48VH-TO-SW06) due to Over Current Protection
Nov 15 22:43:52 switch[45]: Resetting over current protection on port 7
Nov 15 22:43:57 switch[45]: Port 7 cable check results: Open, Open, Open, Open
Nov 15 22:43:57 switch[45]: Port 7 passed PoE Smart cable check, turning on power
Nov 15 22:44:07 switch[45]: Resetting over current protection on port 6
Nov 15 22:44:11 switch[45]: Port 6 cable check results: Open, Open, Open, Open
Nov 15 22:44:11 switch[45]: Port 6 passed PoE Smart cable check, turning on power

[Stephen]

Glad to hear it helped, OCP issue might be a problem with init logic. Will investigate. If you don't mind, can you post your config for the ws3 that is doing this? It may assist with debugging.

[backtick]

Thanks, Sent you the file. I did a default and now the problems have disappeared. Ill keep you posted

edit: After resting for 5min and after cold boot still see the same ports giving OCP error

[Stephen]

Hello backtick,

Been looking into this one. Actually I can't seem to reproduce.
- tried with provided config after modifying slightly (password hash, ip's to work on my network, removed dropbear sshkey, etc) then did coldboot and software reboot
- set all available ports to 24V, 24VH, 48V, 48VH, software reboot and cold boot
- plugged in load within parameters and did coldboot/software boot

I have a few questions that might help us get some answers:
- my current model under test is a WS3-14-600-DC, is yours an AC?
- have these ports been used already and perhaps shorted at any point?
- do you have another WS3 you can check to see if you see the same issue's?

[sirhc]

Thanks, Sent you the file. I did a default and now the problems have disappeared. Ill keep you posted

edit: After resting for 5min and after cold boot still see the same ports giving OCP error

It could be a damaged or defective port. I would suggest RMA

[CircleBWireless]

First off id like to say there is seriously something wrong with this website. It takes 2-3 minutes just to load a simple page. Ive trtied 3 different computers and 2 phones across 4 different internet connections, all results are the same... EXTREMELY LONG LOAD TIMES!

Now, for the issues with the WS3. I have bought 2 of them because i thought the first one was faulty. I had a WS-500-DC that blew up the port board and wouldnt link up at anything besides 10mb on any of the ports. That switch was in service for about 3 years. RSTP worked like a charm on that switch. The original setup was an AF11 link and a AF60XR link with a ubiquiti edgeswitch at one side and the Netonix that blew up.

Bought the WS3 as a replacement due to hard to find inventory at the time. I programed it identical to the previous DC switch and plugged everything in at the tower. After countless hours of troubleshooting and trying to figure out what was going on, i finally decided that the switch was not honoring RSTP. I mean RSTP would NOT work on that switch.....constant loops, wouldnt disable the failover port when the primary was on, itd stay in forwarding 100% of the time. The program on the edgeswitch wasnt changed mind you. After not being able to get the WS3 switch to honor RSTP and confirming that the config was indeed right and setup properly on the edgeswitch via testing 2 other switches i had, i ordered another Ws3. Exact same thing! I didnt copy the config over on the new WS3, I literally hand jammed it all in thinking maybe i missed something in the initial setup. Loaded up this firmware and same ordeal. Loaded up the stable version and same ordeal. Tried this version yet again and same ordeal.

The other 2 switches i had on hand that i did the testing with that both behaved like youd expect regarding RSTP but wouldnt work long term due to needing tons of POE to power all the antennas

A couple weeks ago i was able to locate a WS-26-400-AC model, so i ordered it and hooked it up after programming. RSTP works like a champ, just like you'd expect. I went back and double checked the config on the WS3 against the known working config on the new unit, they're identical. Below is a description of the WS3 RSTP config, pretty basic stuff

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP with port priority of 128 on both and cost on port 22 set to 20000 and port 24 set to 200000

WS3 is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port priority of 128 on both. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

With the above settings failover would NOT work and the WS3 would send packets out of both port 1 and 2 causing a loop. I would have to go into the GUI and disable one of the ports to stop the loop. We use RSTP quite extensively across the network and hadnt ever seen this. Most of the switches that we use are the WISP switch, as well as UISP switches and the edgeswitch when we can get them in, in addition to some FS.com switches. Most of the towers we have are ring topology with dual links, we are migrating to OSPF but these things take time. This section of the network is way older and just has these 2 links anyway, Back to the issues with the WS3.....

In an effort to make the switch work we tried the following settings: WOULD NOT WORK!

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP. Port 22 priority set to 0 port 24 set to 128. Cost on port 22 set to 20000 and port 24 set to 200000

WS3 is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port 1 priority set to 0 port 2 priority set to 128. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

When i received the WS-26-400-AC i programmed it as follows: WORKS LIKE A CHAMP!

WS-26-400-AC is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port priority of 128 on both ports. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP with port priority of 128 on both and cost on port 22 set to 20000 and port 24 set to 200000

This config above worked as it should. Like i said, we used 2 other switches besides the Netonix WS3 while troubleshooting and all of them behaved as expected.

Also, the CONSTANT OCP desperately needs to be fixed. Running 48VH on some LTU equipment itd trip every couple minutes. Heck i even replaced all the wires and 3 of the radios thinking that maybe they were bad, didnt matter OCP kept coming on. Had to run 48v to stop it from coming on. This switch could be really awesome, the DEV time needs to be invested into it though. Such simple things shouldnt be an issue. After purchasing 2 - $1000 paper weights that cant even do RSTP properly, nor keep something powered up without throwing an OCP error, we will be sticking to the normal WISP switch series from here on out. This switch has been out PLENTY long enough that simple bugs like this shouldnt be an issue

[htsinc]

I just upgraded a WS3 to 2.0.9rc1 and defaulted a switch per a conversation on the phone today. I have a very basic config, A management VLAN and a seperate VLAN for a backhaul(AF60XR). I did LAG the fiber ports to a CCR2004. I just tried to power up said AF60XR and immediately get the OCP, using port 5. Here is an excerpt of the log: -- I did try to reset the OCP once. Just using 48V, I guess I could try 48VH...
Nov 16 04:15:53 switch[45]: PoE enabled on port 5, PoE Smart is starting cable check
Nov 16 04:15:58 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:15:58 switch[45]: Port 5 passed PoE Smart cable check, turning on power
Nov 16 04:15:59 switch[45]: Disabling PoE power on port 5 (Port 5) due to Over Current Protection
Nov 16 04:16:04 switch[45]: Resetting over current protection on port 5
Nov 16 04:16:09 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:16:09 switch[45]: Port 5 passed PoE Smart cable check, turning on power
Nov 16 04:16:10 switch[45]: Disabling PoE power on port 5 (Port 5) due to Over Current Protection
Nov 16 10:16:16 UI: Configuration changed by admin (216.198.173.142)
Nov 16 10:16:16 UI: Port 5 PoE: changed from '48V' to 'Off'
Nov 16 04:16:16 switch[45]: Resetting over current protection on port 5
Nov 16 04:16:21 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:16:21 switch[45]: Port 5 passed PoE Smart cable check, turning on power

[sirhc]

I just upgraded a WS3 to 2.0.9rc1 and defaulted a switch per a conversation on the phone today. I have a very basic config, A management VLAN and a seperate VLAN for a backhaul(AF60XR). I did LAG the fiber ports to a CCR2004. I just tried to power up said AF60XR and immediately get the OCP, using port 5. Here is an excerpt of the log: -- I did try to reset the OCP once. Just using 48V, I guess I could try 48VH...
Nov 16 04:15:53 switch[45]: PoE enabled on port 5, PoE Smart is starting cable check
Nov 16 04:15:58 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:15:58 switch[45]: Port 5 passed PoE Smart cable check, turning on power
Nov 16 04:15:59 switch[45]: Disabling PoE power on port 5 (Port 5) due to Over Current Protection
Nov 16 04:16:04 switch[45]: Resetting over current protection on port 5
Nov 16 04:16:09 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:16:09 switch[45]: Port 5 passed PoE Smart cable check, turning on power
Nov 16 04:16:10 switch[45]: Disabling PoE power on port 5 (Port 5) due to Over Current Protection
Nov 16 10:16:16 UI: Configuration changed by admin (216.198.173.142)
Nov 16 10:16:16 UI: Port 5 PoE: changed from '48V' to 'Off'
Nov 16 04:16:16 switch[45]: Resetting over current protection on port 5
Nov 16 04:16:21 switch[45]: Port 5 cable check results: Open, Open, Open, Open
Nov 16 04:16:21 switch[45]: Port 5 passed PoE Smart cable check, turning on power

Try to determine if the port is defective or damaged by using another port.
See if the issue follows the cable or if it is just port 5

[sirhc]

First off id like to say there is seriously something wrong with this website. It takes 2-3 minutes just to load a simple page. Ive trtied 3 different computers and 2 phones across 4 different internet connections, all results are the same... EXTREMELY LONG LOAD TIMES!e

We fixed the forum, spam bots pounding it

F

Now, for the issues with the WS3. I have bought 2 of them because i thought the first one was faulty. I had a WS-500-DC that blew up the port board and wouldnt link up at anything besides 10mb on any of the ports. That switch was in service for about 3 years. RSTP worked like a charm on that switch. The original setup was an AF11 link and a AF60XR link with a ubiquiti edgeswitch at one side and the Netonix that blew up.

Bought the WS3 as a replacement due to hard to find inventory at the time. I programed it identical to the previous DC switch and plugged everything in at the tower. After countless hours of troubleshooting and trying to figure out what was going on, i finally decided that the switch was not honoring RSTP. I mean RSTP would NOT work on that switch.....constant loops, wouldnt disable the failover port when the primary was on, itd stay in forwarding 100% of the time. The program on the edgeswitch wasnt changed mind you. After not being able to get the WS3 switch to honor RSTP and confirming that the config was indeed right and setup properly on the edgeswitch via testing 2 other switches i had, i ordered another Ws3. Exact same thing! I didnt copy the config over on the new WS3, I literally hand jammed it all in thinking maybe i missed something in the initial setup. Loaded up this firmware and same ordeal. Loaded up the stable version and same ordeal. Tried this version yet again and same ordeal.

The other 2 switches i had on hand that i did the testing with that both behaved like youd expect regarding RSTP but wouldnt work long term due to needing tons of POE to power all the antennas

A couple weeks ago i was able to locate a WS-26-400-AC model, so i ordered it and hooked it up after programming. RSTP works like a champ, just like you'd expect. I went back and double checked the config on the WS3 against the known working config on the new unit, they're identical. Below is a description of the WS3 RSTP config, pretty basic stuff

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP with port priority of 128 on both and cost on port 22 set to 20000 and port 24 set to 200000

WS3 is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port priority of 128 on both. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

With the above settings failover would NOT work and the WS3 would send packets out of both port 1 and 2 causing a loop. I would have to go into the GUI and disable one of the ports to stop the loop. We use RSTP quite extensively across the network and hadnt ever seen this. Most of the switches that we use are the WISP switch, as well as UISP switches and the edgeswitch when we can get them in, in addition to some FS.com switches. Most of the towers we have are ring topology with dual links, we are migrating to OSPF but these things take time. This section of the network is way older and just has these 2 links anyway, Back to the issues with the WS3.....

In an effort to make the switch work we tried the following settings: WOULD NOT WORK!

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP. Port 22 priority set to 0 port 24 set to 128. Cost on port 22 set to 20000 and port 24 set to 200000

WS3 is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port 1 priority set to 0 port 2 priority set to 128. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

When i received the WS-26-400-AC i programmed it as follows: WORKS LIKE A CHAMP!

WS-26-400-AC is 8192 - Port 1 (main) and 2 (backup) set to RSTP. Port priority of 128 on both ports. Port 1 (main) set to 20000 and Port 2 (backup) set to 200000

Edgeswitch is 4096 - port 22 (main) & 24 (backup) set to RSTP with port priority of 128 on both and cost on port 22 set to 20000 and port 24 set to 200000

This config above worked as it should. Like i said, we used 2 other switches besides the Netonix WS3 while troubleshooting and all of them behaved as expected.

We have determined STP/RSTP/MSTP is broken - working on a fix.

Also, the CONSTANT OCP desperately needs to be fixed. Running 48VH on some LTU equipment itd trip every couple minutes. Heck i even replaced all the wires and 3 of the radios thinking that maybe they were bad, didnt matter OCP kept coming on. Had to run 48v to stop it from coming on. This switch could be really awesome, the DEV time needs to be invested into it though. Such simple things shouldnt be an issue. After purchasing 2 - $1000 paper weights that cant even do RSTP properly, nor keep something powered up without throwing an OCP error, we will be sticking to the normal WISP switch series from here on out. This switch has been out PLENTY long enough that simple bugs like this shouldnt be an issue

If you have an OCP on a port try using a different port see if bad port or problem follows the cable.

Also another issue just discovered inserting some if not all 1G SFP modules causes a reboot
Inserting 10G SFP+ modules work fine.

Hoping to have a new release in a couple weeks.