v2.0.6 Bug Reports and Comments - WS3 Firmware

[Stephen]

FIXED/CHANGED
- calibration systems and current measurement systems completely rebuilt from scratch to fix measurement issue's after OCP fix (related post)
- reworked boot system to fix occasional reported unpredictable behavior at startup.
- fixed PoE changes on any port would briefly turn off PoE on all active ports
- In RC2, switch had difficulty accessing powered radios
- changing hostname in CLI or webGUI caused random behavior
- Management VLANs with tagged ports can now pass traffic
- updates to software to accommodate new hardware updates to improve OCP functionality
*** IMPORTANT NOTE*** This applies specifically for WS3's rev D and greater see this post
- selectable SFP+ speeds updated to include 1G
- Wattage/PoE popup balloons updating correctly for monitoring
- External Sensor monitoring error when removed from switch

ENHANCEMENTS
- "calibrate-ports" command added to CLI, should not typically be needed as devices are calibrated at factory, but should corruption occur, user can run this command to increase port sensor measurement accuracy, most effective on rev D and above.

WARNING do NOT run this command if cables are plugged into ANY non-SFP ports. Also be sure that POE is NOT active on any port, otherwise calibration will NOT occur for that port.

- "show portcals" command added to CLI, display calibration values for port sensors obtained from calibration routine above or values determined at factory.

KNOWN ISSUES
- WEB UI issues when not at 100% Zoom on browser especially on VLAN TAB
- Some language templates need help - please private message Stephen Copeland to help

Released 9/3/2021

[Garnet]

I hate to be the bearer of bad news but some issues that were present in v2.0.6RC4 appear to still be present in this version. I've written up the steps taken leading up the the errors and attached them as a text file. I've also attached the output of the serial console. As well I have attached an output that occurred spontaneously while I was writing this bug report.

As you can see the issues started when I attempted a config restore.

[Garnet]

Quick update, since I posted that bug report the switch has rebooted itself again, the following was displayed. Appears to be the same error as earlier bug reports.

[Stephen]

Thanks for the report Garnet,
The vast majority of my time for this build was dedicated to calibration routine's for the PoE sensors because we had to change some hardware that required rewriting basically the entire internals regarding the current sensor's - but I had made some other change's that I had hoped would solve some of the issue's you had mentioned earlier.
Either way with this build released, for the WS3 I can focus on the issue's you've mentioned here and our earlier thread.

[gnorton]

Updated from RC5 and still seeing all the previous issues I've reported. Basic config changes like setting an IP, changing the device name or updating the Management VLAN seem to break the switch. It requires defaulting to restore access. I also saw the same PHP errors mentioned above while I had a console cable connected.

[wirelesscon]

Hello regarding WS3 what is the packet Buffer size on the switch this would be important for handling the Speed Difference between the 10G uplinks and the 1G down link Ports

Thanks for your time

[Garnet]

Thanks for the report Garnet,
The vast majority of my time for this build was dedicated to calibration routine's for the PoE sensors because we had to change some hardware that required rewriting basically the entire internals regarding the current sensor's - but I had made some other change's that I had hoped would solve some of the issue's you had mentioned earlier.
Either way with this build released, for the WS3 I can focus on the issue's you've mentioned here and our earlier thread.

Hi Stephen,

It's impossible to say without knowing the original fault trigger but the switch does appear to be more stable. Since my last post I have determined that there is still a fault trigger in the configuration options. It appears that changing IPv4 settings in conjunction with any other setting causes a switch lockup accompanied by the same fwrite and PHP errors, as of the moment the only fix is to factory default the switch as after the problem occurs there is no GUI access.

The issue occurs if:
- IPv4 DHCP is enabled alongside another setting change
- IPv4 DCHP is disabled alongside another setting change
- IPv4 static IP is changed alongside another setting change

In all three cases the switch states the configuration changes was successful but the IPv4 address does not change. If the switch is then rebooted it will experience a lockup lasting approx 10-15 minutes before displaying the same errors as previously mentioned (PHP and fwrite). Other settings changed do appear to take affect (IPv6 settings not tested).

The only other way I've managed to break the switch so far since the new firmware is uploading a config file that was created with 2.0.6RC4.

I hope that narrows down the search for you.

Regards