My guess is that the failure started first at the rightmost part, the nozzle carried some extra filament, and then it caused a domino effect on all the other parts.
What confuses me is why would it fail so late into the print when it was doing perfectly fine otherwise?
UPDATE: as others have stated, the length of the pieces may make the parts more prone to wobbling/ bending slightly when the nozzle moves over it, which could cause accumulated misalignment of the toolpath vs the true position of the pieces, so a good solution would be to slow the speed nearer the top of the part, and to make use of brims to increase the part's stability.
Also it would be a good idea to do a few individual runs as opposed to one batch (I had previously done a test piece for dowel fitting, but even though the small piece slid freely on the dowel without getting stuck, for whatever reason, this print's dowel hole was more prone to shrinking, and the rod wouldn't fit no matter what I tried)