replication: automatic fails, manual "run now" works fine for 1/3 jobs

[artlessknave]

i have 3 replications that are configured the same, they replicate from 3 datasets to 3 datasets (names shortened) using a dedicated user over ssh with keys

lore/bkp - nas1/nas3/bkp/bkp
lore/home - nas1/nas3/lore/data
/lore/data - nas1/nas3/home/home

bkp and home work every time. data fails with every automatic run. manually clicking "run now", however, works every time.
Replication "lore/data - nas3/lore/data" failed: cannot mount 'nas1/nas3/lore/data': failed to create mountpoint: Permission denied
property may be set but unable to remount filesystem.

I could see if both automatic and manual runs fail or all the jobs fail....but having only the automatic fail on 1/3 jobs is stumping me. iv tried staggering timing, ive tried running it with the snapshots and with the custom schedule. lore/data is way larger than the other 2 (they are like megabytes), which is the only difference I can think of.
on the remote system "nas1", these datasets are then replicated to "nas2"
the dedicated user has permissions added for both the remote filesysetm and zfs. this worked in 11.3, so the permissions should be right, but when rebulding the jobs for 12...1/3 fail.

 

[artlessknave]

I figured part of it out. I had a plex jail mounting nas1/nas3/lore/data/Media, which the old replication system didnt care about, but the new one...really doesn't like. changed it to simply mount nas1/nas3/lore/data, and the failures stopped. still doesn't explain to me why automatic failed while manual worked but w/e.
now i just need to fix the plex library folders....but plex has decided Im not allowed to reach any of the settings....
/sigh
good thing I already planned to rebuild the plex jail into a plex plugin.....

 

[artlessknave]

hmm. still fails. sometimes. i feel like im just talking to myself at this point though. :'(