My really dumb tube mistake(s)

Not only should power tubes be replaced for reasons of performance, letting them go too long will cause damage to many amplifiers.  Typically amps with a manual bias.  We output tubes can go into a runaway state where they draw ever more plate currently.  Hopefully just the cathode resistor goes, but in some cases, the output transformer is the victim.  

Being fairly new to tubes and dealing with 2 reputable guys in the NOS tube business, I’ve had very good luck thus far until the other day. I got some 1958 Mullard 12AU7 long plates several months back that I put in my Primaluna Dialogue Premium preamp.
The other day I was showing off my turntable to a friend and I was knocking on the maple isolation shelf while it was playing showing how the table doesnt skip a beat.
It seemed like I could hear a door bell off in the distance. After some investigating found it was tube in my preamp one shelf below. As I tapped or even rubbed a finger on the glass it sounded like a ringing door bell. The tube is now crackling weeks later.
That’s my first experience with a tube going bad.
What I paid for a matched Octet of Mullard EL34 XF2’s has my fingers crossed on those.
Although I have seen several posts here and elsewhere about owners of Primaluna amps using current production Gold Lion KT88’s with favorable results. Maybe a second set of power tubes to switch out now and then would be a good idea to keep the hours down on the Mullards.

Sorry for the long post, but I felt the explanation would help.
This is a subject that is for me difficult to cover without explanation.

I personally look at tubes as an electronic Interface, that has a ease of use, as a method for carrying out exchanges. 

My personal experience from taking part in carrying out exchanges,
has on most occasions been done as part of group, with myself supplying the device to undergo Tube Rolling.  
Doing this has allowed me to experience quite a lot of Tubes, of which some are very well regarded.
I have heard well regarded Valves, that when placed into a Circuit, 
which have almost needed to be removed with immediate effect, 
as the SQ was not one that was wanted.
This was not a personal choice, but a request by a group in attendance.

If I recollect a Orange Print Amperex Bugle Boy was one such Valve used in a DAC,I knew of it, I was interested in it, when used, 
it was a very short lived audition.
On a later occasion, the exact same pair when used on a Phonostage,  were very well received, if a little thin in the presentation.

My experiences allows me to say that Valves with a Good Reputation and are to be found as recommendations, can fail to deliver a good result,
when used in a certain electronic interface.
So jumping into Tube Rolling without acquiring knowledge, seems to be a increased risk, to successfully finding a Personal Sweet Spot with the Valve selection.    

Tube rolling can prove costly, the cost can be impacted by Valve Selection, or accumulating Valves that have not met expectations, and then there are other pitfalls of being supplied Valves that are not measuring up to a usable standard.

I ordered during 2020 Two Matched Pairs of Same Brand 1960's ECC 83 from the same Vendor.
With the intention of producing Two Close Matched Sets.

It was agreed that I would have them Tested on a AVO Test Rig and a return would be accepted pending the test .
One Pair measured very high and slightly over the New Measurement,
(very pleased).
One Pair was between 50% to 70%, depending on the Valve
It took two further supplies of ECC83's that were still measuring quite low, before I got a refund on one pair.
Not all Vendors will be so tolerant.  
Prior to this, a Vendor sent me abusive mails for requesting a Valve return following a Test carried out at my end. 
  
If a modern $20 - 40 Rectifier Valve is claimed to have a similar 
 SQ / Performance of a $200 NOS Valve, and a person is entering into the world of Tube Rolling, the cheaper Valve would be the most cost effective place to start, if a Valve is to be exchanged. 
There should be a change in a performance that will allow an assessment as to the value of the addition.  

I have devices that use ECC88, ECC81, ECC83, 6SN7 and 845's
I have been supported by being loaned Valves that are Closely Matched NOS,  that have been Tube Rolled, those being all the above apart from the 845's. 
As a result some of my Valve Choices have been, more a case of having the Valve to be used already identified, and a purchase tested, to make sure the Valve is sold as described.     

This may seem like a dumb question, but can a tube tester provide the status of a tube prior to failure?