Define too long. 
Because the U.S. Navy had no problem sending us on six-month long overseas deployments. I’ve spent as long as four months continuously submerged. And even between major deployments, we spent more than 80% of our time at sea. The taxpayers aren’t paying for billion-dollar vessels to stay tied up to the pier, after all.
I see, didn’t know that.
For some reason I recalled reading somewhere that the human mental endurance in such confined space was only about 2 months. Guess I underestimated you dolphins. 
He never said he came out of it sane.
The fuel is packaged into the reactor core, so there is no removing and replacing individual fuel elements; at ‘refueling’ the entire core is removed and replaced during mid-life overhaul. It should be understood that the fuel elements in a nuclear reactor aren’t consumed they way a chemical fuel like oil or coal is; they produce energy in which a thermal neutron is absorbed by the enriched uranium (235U) and undergoes fission (~82% of captures) or produces 236U, while the overall volume remains essentially constant and the reactor requires higher neutron flux levels to achieve the same level of output as it ages. In order to make a reactor running at some constant power level last longer, the core either has to be made larger (impractical as the size is limited by the diameter of the pressure hull) or by using a higher level of enrichment in the fuel. There are some other nuances regarding neutron absorption by non-power generating isotopes and neutron damage to materials within the reactor that govern core and reactor lifetime but it isn’t as if the engineering staff is swapping out fuel elements or doing any kind of service within the reactor core; they are monitoring the state of the nuclear core and servicing the outer coolant loop and power generating system that converts the thermal energy into electricity and mechanical power.
Stranger
I’m sure that @robby can speak to this from the standpoint of actual experience but the psychological evaluation of the ability of ‘nucs’ to be able to work in confined situations with a relatively small crew for months on end is as much a part of selecting appropriate people as is mastery of the technical skills. You really have to be able to tolerate the sight, sounds, and especially smells of people you may not particularly like without losing your temper in addition to all of the other inconveniences and indignities of living abort a submarine, so it definitely isn’t for everyone, and of the ‘nucs’ I’ve known most have gotten pretty tired of the life after a few cruises even though they essentially have a multi-month vacation when they return (unless they are schooling for advancement or doing some other posting), and having crew members ‘crack’ under the pressure is not unheard of, although it mostly consists of underperformance or refusal to do duty rather than a Crimson Tide-style conflict. When people talk of sending a crew of four or six people on a multiyear mission to Mars on the basis of experience with 4-6 month submarine deterrence patrols I have to wonder how much they really understand the stress that would entail.
Anyway, getting back to the question of the o.p., here is what an at-sea resupply of a ballistic missile submarine looks like. As a diver, I cannot even imagine trying to do that while submerged. Divers have, of course, operated out of submarines to do things like placing taps on underwater cables or recover equipment, but those require special handling equipment and a dry shelter on the deck for ingress and egress, and the divers spend months training to do particular operations. Moving bulk cargo while submerged would just be an enormous risk for no good reason.
Stranger
On the SmarterEveryDay YouTube channel, Destin has a series of videos in which he spends time aboard a nuclear sub. Rather than electrolysis, at one point he shows the crew burning an “oxygen candle” to adjust the O2 concentration in the sub:
I didn’t watch the video in question, but our normal procedure was to simply bleed oxygen as needed into the ship’s atmosphere from the high-pressure oxygen storage tanks. The electrolytic generators were operated near-continuously to keep the O2 storage tanks topped off. (The only time the generators weren’t operated is for maintenance or drills.)
Fun fact: the oxygen generators used electrolysis to split water molecules into gaseous oxygen and hydrogen. The oxygen was stored while the hydrogen was expelled overboard. So you had gaseous oxygen, hydrogen, and lots and lots of electricity. What could go wrong? 
Did I mention we called the oxygen generator the “bomb”?
Anyway, the only time you’d need to use oxygen candles is if the oxygen generators were down for an extended period of time and the storage tanks were getting low. The candles were basically emergency equipment. Similarly, we had carbon dioxide absorbent consisting canisters of lithium hydroxide that was also for emergency use.
It could be that a submarine crew might use oxygen candles for practice, or to rotate stock, or maybe their oxygen generator was down for an extended period of time, but I don’t recall ever using them.
AIUI, some of the Kursk submariners kept burning oxygen candles to survive as their compartment was half-flooded with water. Eventually they suffocated anyway.
Actually, the oxygen generating cartridges probably brought a swift end to the survivors in the ninth compartment:
From Wikipedia:
Or you can do what the San Juan did [twice] - hijack the supply run of an arctic expedition, then go back and take their resupply when it arrived … not quite doing it underwater, but resupply without actually going into port =)
I was curious as to why they would have died of carbon monoxide poisoning and not simply asphyxiation (i.e., lack of oxygen in the breathable air). After browsing Wikipedia for a while it turns out that asphyxiation kills you in about seven minutes, whereas breathing a 1.28% concentration of carbon monoxide kills you in less than three.
If the inhaled gas is truly oxygen-free, then victims are rendered unconscious after just a few breaths, same as breathing normal air with 1.28% CO.
Yes, that’s the same article I found, and the one which states that death occurs after seven minutes. Exactly when consciousness is lost isn’t particularly relevant to the claim about what killed the Kursk submariners.
The Russians being secretive especially about disasters that were an embarrassment to Putin haven’t fully shared a lot of the details of that disaster, but I also think in some of the compartments they explored during the recovery it was found the CO2 levels were higher than they would have been just from human respiration, suggesting something was burning etc as well.