Solved: Re: Catalyst 3850 sho env power produces odd results

brianw360 · ‎08-25-2022

Here's the setup:

Switch Ports Model SW Version SW Image Mode
------ ----- ----- ---------- ---------- ----
* 1 16 WS-C3850-12XS 03.07.04E cat3k_caa-universalk9 INSTALL
2 32 WS-C3850-24S 03.07.04E cat3k_caa-universalk9 INSTALL

There is a stack power cable being used but not two. The left power slot ONLY in both units have power. There was a PS in the right power slot but power cord not attached.

Replaced the PS in the right slot with a known good and still getting this response, which is the same before I replaced it.

WAU-MIS-CORE1#sho env power
SW PID Serial# Status Sys Pwr PoE Pwr Watts
-- ------------------ ---------- --------------- ------- ------- -----
1A PWR-C1-350WAC LIT21164DXZ OK Good Good 350
1B Unknown Unknown No Input Power Bad Bad 235

How is "1A" and "1B" being used in this reference. My reasoning, it would be for switch one in the stack. But it makes sense to me that since there are not two stack data cables that this could be the reason for the "Unknown Bad" comments for 1B.

Thank you for your time.

Brian

Leo Laohoo · ‎08-25-2022

The command "sh env power" is a bug in itself. It will fool the operator to think there is only one switch. The correct command is "sh env alarm" and this will enumerate every PS in the stack. All of them.

Going back to the output, if someone is at the back of the switch and looking at the two power supplies, the inner-most power supply is 1A. The outer-most PS is 1B.

Finally, the output will depend on the firmware version. There are bugs in the 3.X.X where the PS is working fine but the switch is reporting them as faulty or failed. Before going out to inspect/replace the PS, reboot the stack and see if the PS status looks better.

A Word of WARNING: Do not make any attempts to upgrade the firmware to 16.X.X. Stay at 3.X.X.

View solution in original post

Leo Laohoo · ‎08-25-2022

The command "sh env power" is a bug in itself. It will fool the operator to think there is only one switch. The correct command is "sh env alarm" and this will enumerate every PS in the stack. All of them.

Going back to the output, if someone is at the back of the switch and looking at the two power supplies, the inner-most power supply is 1A. The outer-most PS is 1B.

Finally, the output will depend on the firmware version. There are bugs in the 3.X.X where the PS is working fine but the switch is reporting them as faulty or failed. Before going out to inspect/replace the PS, reboot the stack and see if the PS status looks better.

A Word of WARNING: Do not make any attempts to upgrade the firmware to 16.X.X. Stay at 3.X.X.

brianw360 · ‎09-20-2022

Leo...can you expand on why you do not recommend upgrading from 3.7.4 to 16.12.x??

Thank you Sir!

Leo Laohoo · ‎09-20-2022

@brianw360 wrote:
can you expand on why you do not recommend upgrading from 3.7.4 to 16.12.x??

Because IOS-XE leaks like a sieve.

I have several stacks of 3850. A small amount of them are still hanging on to 3.X.X and the rest are on 16.X.X. Guess which ones have an uptime of >4 years -- And not even a single Traceback.

I would never (ever!) "recommend" anyone (still) on 3.X.X to upgrade to 16.X.X.

brianw360 · ‎09-21-2022

Thank you Sir. Wow what a write up!

brianw360 · ‎08-27-2022

Thank you Leo for your mentoring. Very much appreciated.

These old switches are on an OT network and luckily we have a planned outage at the mill next month. I've seen many switches that have be online exceeding years.