Yeah on CWA/CWI ballasts, a shorted capacitor can cause the ballast to draw excess current and burn it out if left unchecked. I guess you could also monitor input current with a clamp meter or a kill a watt and compare to nameplate current.

When I test a new to me used HID light, especially an older one I monitor the current with a kill a watt meter and I also probe the ballast core temperature with a IR temp gun from time to time.

Unfortunately while it's nice to keep a streetlight all original, I treat capacitors as a consumable item like a lamp. My personal compromise is replacing oil filled with oil filled instead of the dry film plastic type used in modern HID lights.

For obscure hardware and fasteners, I generally order those from McMaster-Carr. They have a decent selection of fasteners and their website is laid out well.

Yeah on CWA/CWI ballasts, a shorted capacitor can cause the ballast to draw excess current and burn it out if left unchecked. I guess you could also monitor input current with a clamp meter or a kill a watt and compare to nameplate current.

When I test a new to me used HID light, especially an older one I monitor the current with a kill a watt meter and I also probe the ballast core temperature with a IR temp gun from time to time.

Unfortunately while it's nice to keep a streetlight all original, I treat capacitors as a consumable item like a lamp. My personal compromise is replacing oil filled with oil filled instead of the dry film plastic type used in modern HID lights.

For obscure hardware and fasteners, I generally order those from McMaster-Carr. They have a decent selection of fasteners and their website is laid out well.

I might look into buying some testers so I can check things like that, as well as a temperature gun.

I'll check that website for the slipfitter bolts. I might need to do some research as I'm not a expert on hardware. I'll need help figuring out the threads for sure. Size I can probably pretty easily figure out. I just want bolts that look the same as the original. The stainless, or zinc steel at Home Depot or Lowes are the wrong color. Silver hardware would look off.

I don't even have any instruments for testing amp draw, capacitance, and things like that. I'd have to buy one. First I'd have to know what I'm looking for when I go to the store lol.

The capacitance meter is a convenience when you have to deal with capacitors loosing their sapacitance, but you have to be familiar how it responds to other failures like elevated leakage or ESR or similar. Most of them tend to severely mislead you, if you do not check or count for it(in case of an intentional integrated safety bleeder resistor).
But it is not 100% necessary, as you may test the capacitors, even in a more reliable way (because being in direct control of the test method used).

So for lighting I would rate the instruments in following way (from the most important till the ones just "nice to have"):
- AC and DC voltmeter, rated at least 400VAC/1kVDC
- Continuity tester with beeper
- Ohm meter capable measuring down to 0.1 Ohm accuracy at resistances in the Ohms range, up to 10 MOhm
- AC A-meter, best a clamp meter
- Isolation quality tester (aka "MegMet" or similar names; 1kV test voltage should be far enough)
- "Cheat sheet" with various electrical equations, a calculator, stack os scratch papers and a few writing tools
- Real AC power measuring device (kill-a-watt,...)
- DC A-meter
- current limitting power source (a board with switchable incandescents of various power rating in series with the output; aka "Dim Bulb Tester", google what it means)
- isolating transformer (feeding your test socket)
- an oscilloscope (isolating transformer mentioned above is a must then!) And learn basics how to use it. Even 1MHz is enough (for mains frequenncy ballasts; 100MHz for electronic ones), but be sure you have 600V rated test probes for it.
- current probes for the oscilloscope (could be hacked out from some cheap clamp meter - add a signal output to it)
- for electronic ballasts:
  - Diode test (uses to be part of common multimeters)
  - ESR meter
- Only then are the various C and L meters

Like Ryan said above, I don't want PCB capacitors indoors where they can leak or smoke when they fail. I've had this happen before, and the smell is hard to forget.

Yep, had a PCB capacitor go pop in a slimline ballast, had to replace the whole fixture because it smelled awful and the cover had a pool of oil that spilled when i opened the fixture up to look at the damage (which was fun to clean).

For film capacitors (pretty much all modern lighting capacitors...) they are all of the self healing type, so unless there are serious manufacturing quality problems with the capacitor, it will start losing capacitance long before other types of failures happen

Older capacitors can be paper with foil or metallised paper. Those could have an internal destruction process going on without noticable change to capacitance. And it ill appear as the capacitor getting hot. But such processes start and stop abruptly, and without the self healing mechanism of metallised film capacitors, there is nothing in place to stop it - even the first time. The capacitor may be good, then after a while go into breakdown and explode in a few seconds to few minutes

(Modern metal can capacitors have a safety cutoff that's based on the can expanding and tearing the internal connection, but i guess capacitors from the PCB era didn't have such mechanism)

Anyway, i find it quite unreasonable that good capacitors would heat much. Film capacitors are fairly lowloss, the power they are supposed to dissipate is quite small  vs. the capacitor size (so surface area for cooling)



Besides

A couple years back i found 5 Fluorescent luminaires from the early 70s with PF capacitors. (Elco capacitors, based on Fribourg capacitors licensed designs. Those were made here in the 60s/70s). Those capacitors dont have any safety mechanism in them. Just the capacitor (supposedly oil filled) sealed in an Aluminum can with potted top

While the capacitors look ok, the fittings are covered with some Red-ish residue throughout the insides, which looks like it could well be something that leaked out of the capacitors slowly over the years ?