WS3-14-600-AC Should We RMA?

[Garnet]

Hello,

I'm hoping Stephen or Dave can weigh in on this. I've had a WS3 sitting on my test bench since May and in that time I have been unable to find a setup that doesn't cause frequent crashes and or weird behaviour. This unit has been RMA'd before (in April) and I can provide the original order, shipping, and RMA numbers if that would help. Since we have been unable to get this switch to the point of considering using it in production I would like to know the following:
1. Is our switch a special case, has no-one else had any of these issues?
2. At this point should we just RMA the switch?
3. Seeing as this switch was already RMA'd what would be the timeline on getting a replacement unit (i.e completely new switch)?

[michaeln416]

We are still waiting for a reply to this post. Please and thank you.

[sirhc]

This is not seen anywhere no, not to my knowledge.

Try this

Upgrade to v2.0.7
Factory default
Connect via console cable or via SFP port (no other cables in ports 1-12)
Run the command calibrate from the CLI either serial connection ot via the UI console which will calibrate the current sensors.

You NEED to not have cables in ports 1-12 so it can apply power and calibrate sensors.

Re-set back up manually, DO NOT export and import a configuration as you problem "could" be a corrupted configuration.

If your still having issues RMA the unit.

We are "expecting" some WS3 boards soon.

With the semi-conductor shortages we are having a hard time building switches which is apparent as we are out of stock of most things on our site. - Lets go Brandon!

[Garnet]

We have performed the above procedure and unfortunately problems persist in fact the existing problems with the switch almost prevented us from completing the above procedure. My diagnostic notes are below:

Begin Diagnostic Notes:
- Lots of lockup before test could begin
1. Upgraded WS3 to FW 2.0.7
- Upgrade appeared successful but switch locked up during reboot
2. Defaulted via GUI
- switch booted
4. Ran calibrate
- This is not a valid command
5. Ran calibrate-ports instead
- Ran successfully
6. Re-enabled DHCP via CLI
- success
7. Connected eth3 while entering an 'ip a' command
- success but switch was slow to respond
8. Enabled discovery (all options)
- Appears successful
9. Rebooted
- Switch booted successfully but continues to count down by 1 every time it grabs a new address (e.g. was 132 now is 131)
10. Changed switch name to HT-WS3 and changed login credential name.
- Error saving config
11. Refreshed browser pressed config save button again
- no apparent affect, apply pop-up appeared to just wait for timer to run out
12. Refreshed browser rebooted switch via GUI
- switch lockup

End Diagnostic Notes

From the above we have concluded that this WS3 is in need of a second RMA. On a final note I would like to once again highlight that the trigger for the lockups we were seeing was tied to changing more than one config option at once between reboots. Although there were also instances where the switch would lockup if only one config option was changed.

Previous threads discussing issues with this WS3 are linked below:
First Thread:
viewtopic.php?f=17&t=6948&hilit=can%27t+log+in#p34578

Second Thread:
viewtopic.php?f=17&t=7045&p=35773#p35773

Third Thread:
viewtopic.php?f=17&t=7231&p=35774#p35774

[sirhc]

6. Re-enabled DHCP via CLI -
- success

If I am understanding you, you have the switch getting a DHCP address instead of a static address?

If so please try your testing only using a statically assigned IP address.

7. Connected eth3 while entering an 'ip a' command
- success but switch was slow to respond

Not sure I follow this?

8. Enabled discovery (all options)
- Appears successful

Try not enabling all options and see what happens.

I am asking you to eliminate the above 3 steps and see if you still have issues.

[Garnet]

Hi Chris,

If you read the prior threads it should be evident that this WS3 has some glaring issues. As I mention above we almost couldn't perform your test due to the issues with this switch as it routinely locks up for around 10 - 15 minutes before sending a wall of error messages over the serial connection. This switch has been tested in every conceivable way over the past six months including all ways you and your staff a have suggested. It is obviously a undetected bug in the firmware or faulty hardware.

We putting in an RMA for this switch as whatever was done the first time it was RMA'd did not fix the issues it has (hence why we want an entirely new switch to replace it).

On a final note I'd love to eliminate the three steps you've pointed out but that would mean not testing the switches stability as well it is not relevant what configuration changes are made (as you would know from prior threads) as this WS3 is so unstable that it seems looking at it the wrong way causes the issue.