3nprob

If I get this right, addressing this for WL should solve #3675 / #3697? Considering those, would it make sense to bump doing so?

I also get forced to reaauth every ~1-5 minutes - which is super frustrating obviously. There's nothing exceptional in logs and the backend process is uninterrupted. This is for any...

Got it - was not confident if this was intended behavior or not and it does fit the title (: (I'm on 0.10.0 BTW)

Getting this consistently as well. Receiver end on a single encrypted pull job eventually gets killed after ~500 steps out of 5000 on a 4GB system. Adding virtual memory it...

@problame No no, just added and enabled a 7GB swap file (:

Every other time or so it gets to the point that the whole system hangs to where I have to do a hard reboot (it doesn't freze fully but almost;...

Sorry for not following up on this - I managed to eat through the backlog of snapshots through a series of restarts over several days and don't have logs retained...

So as syncing has resumed, so has this issue (though now it's a matter of hours, not days). Hopefully I can get some more helpful profiling data to help pin...

I think my perspective is a somehwat common one. I run zrepl as a systemd service (using the default resulting from the deb package). I configure all jobs through ansible....

You can already kind of achieve the first option by splitting it up into two jobs with different prefixes. However, I'd also be in favor for (some variation of) the...