Need to go back to Londonium again, due to yet another shitty, crappy, unreliable Cisco server PSU failing. 3rd this week.
Cisco Server? One of those rebadged Dell's that they rebadge and tell you it's an Appliance?
As far as Cisco routers (*) and switches go, I could count the amount of switches and routers i've replaced due to failed PSU's in the last 25 or so years on one hand and i'd say pretty much all of them failed because they got too hot (which is always down to an installation issue - poor airflow in cabinet / comms room - New builds hate installing AC as it messes with the Green credentials of the build). I can't say the same for the total POS HPE Aruba and HPE Comware switches in the past 5 years though.... Thank God the M versions were bought with dual PSU's....
....Still, keeps me employed and paid so I guess every cloud has a silver lining! 
* OK i'll admit the C891's can be a bit shitty if they build up any static
and the 6509's can sometimes get their knickers in a knot when the Supervisor board wont release the power to other boards in the chassis.
They do their own now, after rebadging HP Proliants and then Dell Poweredge. And they are monumentally shite with a massive failure rate (like >20% a year).
And, I discovered this week, if you use the official cable management arm, you can't replace one of the hot plug PSUs without disconnecting the cable management arm, thus unplugging the other PSU. Talk about not thinking that one through.
Generally, I've found their switches and routers to be pretty reliable, including our creaky 6509s

. Although one of our Nexus 5000s ate 3 fans in a fortnight a few weeks ago.
We went through a stage of 40Gb SFP's failing, both in the Cisco servers and in the Nexus. But hardly surprising, as they are too hot to handle after they have been inserted for more than a day or so. No wonder Cisco put those rubber handles on
