v2.0.8 Bug Reports and Comments - WS3 Firmware

[sirhc]

FIXED/CHANGED
- Orphaned terminal sessions that would cause 100% CPU issue
- Issues with SFP+ configuration
- Memory leak that caused reboots
- Security Issues
- SNMP Issues

ENHANCEMENTS
- Added checks that config changes save correctly before applying
- Automatically reload UI if spurious error occurs

NOTE:
There is a known issue with Chrome that can cause requests to randomly cause an error when accessing a site with a self signed certificate.

https://bugs.chromium.org/p/chromium/is ... ?id=488043

We added some bulletproofing to try and handle this, but it could still be an issue.

Potential workarounds:
- Use Edge, Firefox, or some other browser (bonus: Firefox will save the password whereas Chrome won't)

- Enable HTTP (Device -> Configuration -> HTTPS -> Allow HTTP) and use that. Potential security issue, but if you are using an internal management network to admin your devices it probably isn't a problem.

KNOWN ISSUES
- DO NOT USE DASHES IN SWICH NAME CAUSES ISSUES WHEN SAVE OF CONFIG
- WEB UI issues when not at 100% Zoom on browser especially on VLAN TAB
- Some language templates need help

Released 3/11/2023

[michaeln416]

When I upgraded to 2.0.8 from 2.0.8RC2-202212051326, the device did not successfully complete the automatic reboot process after the new firmware was applied. It remained with all ports turned off for 15 minutes, after which I unplugged the power cable and forced another reboot. It appears to be fine after that.

Therefore, for anyone else upgrading, ensure you have someone on-site; don't do this remotely.

[sirbryan]

On 2.0.8 I'm still seeing the same problem as mentioned in this thread (Re: v2.0.6 Bug Reports and Comments - WS3 Firmware) where, if I change the IP address and management VLAN at the same time, I lose access to the switch and have to reset everything to defaults.

I did manage get it into a state one time where it would pick up an IP address via DHCP, respond via ARP and ICMP, but it was rejecting all forms of communication (except serial port).

[turbojack]

I have run into a couple of issues with my first WS3 running v2.0.8. Here is what I have noticed.

1) When making config changes I typically do not get past the 300 second countdown timer. For instance, I can change the switch name and hit apply. The count down time (300 seconds) starts but will never switch back for confirmation of changes even though no ip changes were made. After about 15-20 seconds I can refresh, then I do get confirmation that the changes were successful. Viewing from the console seems to validate that the changes did indeed apply.

2) I have an issue where I changed the switch name from the default to location-sw-01. After this specific change, when cold booting the switch, it takes approximately 10-15 minutes to boot. During boot, the console stops at a specific spot with no errors, just blank. It also will not respond to input. During this time the fans run at 100%. After the 10-15 minute period, the fans make a small slow down blip sound and then the switch console becomes responsive again, but usually the ip address is no longer reachable. I can factory reset from the console or the reset button and then the switch acts normal again. I finally removed the dashes out of the name. I changed it from location-sw-01 to locationsw01 and it appears to function ok without the 10-15 minute boot times. Now, it will sit at the same spot for around 60 seconds then proceed to be an operation switch. Additionally, the ip addressing seems to function properly at that point.

Thanks

[backtick]

Enabling a D vlan on any port seems to lock up the WS3. I've tried twice and locked up both times.

[Garnet]

I have run into a couple of issues with my first WS3 running v2.0.8. Here is what I have noticed.

2) I have an issue where I changed the switch name from the default to location-sw-01. After this specific change, when cold booting the switch, it takes approximately 10-15 minutes to boot. I finally removed the dashes out of the name. I changed it from location-sw-01 to locationsw01 and it appears to function ok without the 10-15 minute boot times.

This sounds like the issue I found and reported about a year ago now on a previous version. A seemingly unprompted switch lockup on every reboot that turned out to be caused by nothing more than dashes in the name.

[dbnetwork]

we find a problem in displaying the power supplied on the individual ports and in the total, below are the photos:

FIXED/CHANGED
- Orphaned terminal sessions that would cause 100% CPU issue
- Issues with SFP+ configuration
- Memory leak that caused reboots
- Security Issues
- SNMP Issues

ENHANCEMENTS
- Added checks that config changes save correctly before applying
- Automatically reload UI if spurious error occurs

NOTE:
There is a known issue with Chrome that can cause requests to randomly cause an error when accessing a site with a self signed certificate.

https://bugs.chromium.org/p/chromium/is ... ?id=488043

We added some bulletproofing to try and handle this, but it could still be an issue.

Potential workarounds:
- Use Edge, Firefox, or some other browser (bonus: Firefox will save the password whereas Chrome won't)

- Enable HTTP (Device -> Configuration -> HTTPS -> Allow HTTP) and use that. Potential security issue, but if you are using an internal management network to admin your devices it probably isn't a problem.

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

Released 3/11/2023

[sirhc]

- DO NOT USE DASHES IN SWICH NAME CAUSES ISSUES WHEN SAVE OF CONFIG

Will be fixed soon in RC release

[Stephen]

RC release is available, please download firmware and report feedback in the v2.0.9rc1 Bug Reports and Comments - WS3 Firmware thread.

[bpeach]

Long time user of Netonix WISP switches here, just installed my first WS3-14-600-DC running stable release 2.0.8. My configuration is fairly simple with only four VLANS, no LAGs, and STP is turned off. Connected to CCR2116 router via 10G. No hyphens in the switch name. Overall we love the new hardware so far, but I do have a couple minor issues to report:

1) I observe the same POE power utilization display issue as previously reported here: Netonix Forums • View topic - v2.0.8 Bug Reports and Comments - WS3 Firmware Further to the OP's report, I believe this erroneous data is also causing inaccurate power supply efficiency data to be reported. I see that the release notes were quoted in response to the original issue report, but I don't see any specific correlation with the issues identified in the release notes.

2) With approximately 300 to 600 Mbps of actual traffic transiting the switch (in both directions), the switch occasionally (every minute or so) reports well over 10Gbps of total traffic and the CPU utilization spikes to ~90% for a few seconds. Often times the GUI temporarily stops updating when this happens. Nothing of note in the logs and the Mikrotik CCR2116 does not report the same traffic. I have loop protection on. I don't really believe the traffic spikes are actually happening (GUI artifact?) but I have not done any packet sniffing or other diagnostics to prove it yet; I thought I'd check to see if anyone else has reported similar issues first. I tried to paste screenshots but unable due to exceeded max allowed characters....

Since I can't paste graphics, here are some observations:
The total traffic graph is peaking at ~15Gbps
The 10G interface traffic graph is peaking at ~6Gbps
The 1G interface traffic graph reporting close to 5Gbps peaks.
The traffic graph numbers look correct and do not match the graphs. Only the graphs themselves seem to be in error.
Tik router does not show any spikes in traffic rate.


Regarding both issues described above, I have tried Firefox, Edge, Safari, and Chrome. Also, I have HTTP enabled.

Any support/advice/comments would be appreciated!