Fairly new player here, playing on monarch/huge/terra. I'm in the late medieval period, and having trouble breaking through longbowmen in hilled cities. I have crusaders, foot knights, and man-at-arms with city attack upgrades, and knights with retreat upgrades, but none of these seem to deal any significant damage to the longbows (even with city defenses at 0%).
That is a relatable feeling!
They're almost impossible to root out of cities in the medieval era pre-gunpowder, especially (as ThirdOrbital said, if they are receiving extraneous bonuses like being on a hill, protective, etc.) It really doesn't show through conspicuously on paper just how effective they are as city defenders, especially when the era is otherwise replete with so many cool and various other units that look powerful and tempting to use as invaders.
Is this just a feature of the era, that longbowmen are exceptionally good at defending cities?
Yes lol, and this is one reason why they're one of a few unit classes that not every civ gets. (All of the unit rosters are nominally supposed to be balanced, but longbows on their own are certainly strongly biased towards the defensive.) I even made the case that they were OP in the pre-3.6 subversion and they were consequently nerfed to become more expensive as of the current version.
Do I just need more units? (I have roughly 10 city attack units to their 5 longbowmen, but again, I don't seem to deal any significant damage regardless.)
That does sound very insufficient unfortunately. A 2:1 ratio here even with premium city attackers is very unlikely to cut it, unless it's a flatland city or they're injured or something. The above suggestions look sound to me, though your best option is most likely to wait until another era if you're planning conquest. If you do find yourself in a situation where some other factor compels you to conquer longbow defended cities in the medieval era, I would personally go with something like a two-pronged attack: I'd build a huge stack of poor quality units (mostly levies, which aren't altogether bad) next to a medium one with premium city attackers, and attack with the former in spite of poor odds just to soften them some before following through with the good units when your odds won't be likely non-existent. The caveat here is that there is obviously a big sunk-cost factor at play and you will want to make sure that you do commit to fighting completely with the second one, since you will end up feeding them an enormous amount of XP with your initial blow otherwise, since you intended to lose the first stack entirely, meaning more promotions to the longbows and the opportunity to heal and become even harder to deal with later while losing your own investment if you don't follow through same turn.
It is important to note, though, that gunpowder itself is a medieval technology, and bombards confer a lot of immediate utility in that they are the first standard unit to inflict cost-free damage from ranged attacks. Even just a little bit of strength reduction can tip the scale of combat odds quite significantly, and take them from nearly untouchable to something you can overcome much more feasibly if albeit with difficulty.
Or do I need higher tech to break through?
As you keep playing and get a feel for the tech tree and unit rosters, you'll see an ebb and flow of what sort of warfare is most effective at any given time. In the medieval era, conquest itself favors the defensive but becomes a lot more volatile in the field. Longbows are still somewhat meaningfully effective against even line infantry (especially well-promoted ones), but only as city defenders. Their counter in the renaissance (the light infantry) does get a malus against city attack, the only useful role of the longbow that late in the game.
Shocking news - at least for me 
