Cisco 3850-48F-L, Denali IOS 16.3.5b, uptime is 9 weeks, 6 days.
Seeing repeating log errors:
Sep 4 12:00:08: %SIF_MGR-1-FAULTY_CABLE:Switch 1 R0/0: sif_mgr: High hardware interrupt seen on switch 1
Sep 4 12:01:08: %SIF_MGR-1-FAULTY_CABLE:Switch 1 R0/0: sif_mgr: High hardware interrupt seen on switch 1
Sep 4 12:01:23: %SIF_MGR-1-FAULTY_CABLE:Switch 2 R0/0: sif_mgr: High hardware interrupt seen on switch 2
Sep 4 12:02:08: %SIF_MGR-1-FAULTY_CABLE:Switch 1 R0/0: sif_mgr: High hardware interrupt seen on switch 1
Any update regarding the problem? I am experiencing the same problem with the same firmware version. In my case, one of the stack member restarted unexpectedly.
Did you restart the switch? Mine already restarted by itself (I guess this counts as restarting the switch) and the problem still persists.
I opened a TAC case via my vendor. The TAC said "try to replace the suspected faulty stack cable if error message still persist". I did not find error messages anymore on the switch log, so I did not proceed to replace my suspected faulty cable. I will still monitor the condition though.
I think there is no downside on trying Leo's suggestion, since the current recommended IOS is 16.3.6.
Please check the manufacturing date of the stacking cables. We have had switch stack reboot with this
12966: Oct 16 12:38:36: %SIF_MGR-1-FAULTY_CABLE:Switch 3 R0/0: sif_mgr: High hardware interrupt seen on switch 3 12967: Oct 16 12:38:37: %SIF_MGR-1-FAULTY_CABLE:Switch 4 R0/0: sif_mgr: High hardware interrupt seen on switch 4
Every time we got this it was a cable with a manufacturing date of 2015. We had 12 cables go bad. It got worse when we upgraded to 16.3.6
We are in the process of finding and changing all the 2015 cables. I'm in an active TAC case... no help yet
om, thanks for the tip, I will try this. I never noticed the date stamp on the stack cables before. An interesting side note is that running the command "show switch stack-ports summary" shows no incrementing changes to link. The cryptic error message I gave in this post's title does not indicate which stack cable(s) is a fault; this is a stack of four 3850-48's. I guess I just have to visually check each cable's sticker.
That reminded me about this.... doing a Show Switch-Ports Sum showed a cable with unknown "Cable Length. This could help you identify a bad cable:
Sw#/Port# Port Status Neighbor Cable Length Link OK Link Active Syn
1/1 OK 3 50cm Yes Yes Yes
1/2 OK 2 50cm Yes Yes Yes
2/1 OK 4 50cm Yes Yes Yes
2/2 OK 1 50cm Yes Yes Yes
3/1 OK 1 50cm Yes Yes Yes
3/2 OK 5 50cm Yes Yes Yes
4/1 OK 2 50cm Yes Yes Yes
4/2 OK 6 50cm Yes Yes Yes
5/1 OK 7 50cm Yes Yes Yes
5/2 OK 3 50cm Yes Yes Yes
6/1 OK 8 50cm Yes Yes Yes
6/2 OK 4 50cm Yes Yes Yes
7/1 OK 5 50cm Yes Yes Yes
7/2 OK 8 Unknown Yes Yes Yes
8/1 OK 6 50cm Yes Yes Yes
8/2 OK 7 Unknown Yes Yes Yes
I also encountered the faulty cable message in my logs after re-installing a stack of 3 3650 48 port switches.
In IOS-XE 16.9.3, the command to check the details of the stack ports is:
show switch stack-ports summary
Curiously, everything looked fine, but the error messages were still being logged. Then after 6 hours the error messages stopped. The stack has been running without complaining about Faulty Cable errors for a week, so maybe it was just an initial hiccup ¯\_(ツ)_/¯