glowing red output tube, but don't think it's red-plating

Usually in an amp that was functioning the main reason is a loss of bias, which is to say a loss of control Voltage on the control grid of the tube. A typical large pentode for example might have a bias Voltage on the grid that is -40 Volts or so with respect to the cathode of that tube. If the grid resistor opened up, or the contact of the tube or the tube socket failed, the control Voltage would be absent and so the tube would conduct as hard as it could- hence red-plating. IOW this is a loss of bias. 

Thank you for the continuing education you are providing, @atmasphere .

The lucky bit is tubes are pretty forgiving about this. If you detect a tube is red plating and shut things down quickly enough, the tube may well survive undamaged, assuming it wasn't the problem in the first place. 

I think I got from the OP that his issue was fixed by replacing the tube, and the system sounds good again.  Meaning that it was a tube failure, which I would think would be preferable to issue with bias due to something happening with the gear. Earlier I referred to a pair of ARCs I used to own that sounded great, but were unreliable.  Often enough that it was not a blue-moon even, on start up one or the other would blow a grid resistor and that tube (it used 6550s) would take on a strange glow.  To be on the safe side, I always swapped out the pair (or sometimes the whole quad) of tubes.  I am assuming that these failures were due to the component and not the tube, although the component malfeasance was bringing down a tube when it occurred?

@xenolith -

Following you post with some interest because you said you had other tube failures. Were they each KT77s? FWIW, I was using Genalex KT77s in a Cayin 50T integrated each of the 77s blew. Kevin @ Upscale said that while they sound great, they are not as rugged/dependable as EL34s.

I've lived through a red-plating event and the screening resistor blowing up in a spectacular fashion. Indoor fireworks isn't fun.

From my 1-time experience and the successor investigation, repair(s), re-tubing, etc. here's my opinion:

Red Plating often involves pulsating. This is not necessarily exclusive to bias, and can be a problem with the tubes that govern the output tubes. 

I'm in the camp that you have a bad tube. FWIW I learned the hard way that all new matched tubes is the preferred and safest method. long story, but in my case a painfully expensive set of matched ARC tubes turned out to be the end of all my amp's problems, and a surprisingly improved sound and volume capability. 

@wharfy the amps are custom made; the only two on earth.  The original designer and person who initiated their build was Gary Dodd.  He died before finishing them.  6 years later in 2021, his friend and colleague Charlie Cocci completed them and sold them to me.  These amps are optimized to run KT77s in 43% tap ultralinear mode; so much so that they literally cannot run any other tube type.  So yes, all previous failures where of (Gold Lion) KT77s.  From the get go, I was able to set the individual bias for each output tube to just barely 40 mA...usually with the bias pot turned all the way down.  Sometimes even at fully turned down the tube would be at ~45 mA.  After about half a dozen spectacular tube failures which always took out the sacrificial resistor on the bias circuit, the fuse and sucked in the glass envelope to the point of breaking,  I contacted Charlie to ask what the heck was going on with these amps.  Long story short ended with him incredulous that the bias circuit could be the problem but that he ultimately was even more incredulous that the three multimeters that measured >40 mA at full down turn of the bias pots could be wrong.  So he advised me to buy two sets of resistors, I forget the values.  Then one day, while he was on speaker phone I turned one of the amps on that had one of the sets of resistors installed...still got 40-ish mA at full down turn, so he told me to swap those resistors with the 2nd set and try it again.  Perfect!  At full turn down the output tubes are at 0 mA and 40 mA is at about 5 turns of the 10 turn Bourns bias pots.  There hasn’t been any tube failure since then until a few nights ago when the I guess red-plating failure occurred.  Much less spectacular and much less hassle as I didn’t have to unplug everything and lift an turn over the 75 lb. beast in order to replace the sacrificial resistor.  Soooo, now that the bias circuits are correct, I’m finding great longevity, which I think can also be considered durability, from the Gold Lion KT77.

I'm convinced that the tube just died of old age.

@xenolith -

Interesting story about the resistors and biasing. Oh the stuff we do, and put up with, in pursuit of the perfect sound.