oom error message too confusing (failed to fill whole buffer)

[lorem111]

internal error: failed to get or launch module host: failed to fill whole buffer

3blave says: "That does seem like it might be an OOM error, not an awesome error message though"

made a ticket because he told me to :)

i put image of what it looks like. cant provide steps to recreate probably i did something stupid and blew up my database. thank you and hope you can baby-proof this <3

[bfops]

Thanks for reporting! We'll try to reproduce on our end, but please let us know if you figure out how to make this happen again 🙂

[egormanga]

Just happened to me as well. I suppose it's been caused by the lack of free disk space for me, or perhaps by a host crash (both happened around the same time).

[egormanga]

The error is:

Error: failed to get or launch module host: failed to init replica 2000001 for <identity>: failed to fill whole buffer

There's an unusual picture in the data structure:

/stdb/data# ls -l replicas/2000001/clog/
total 8280
-rw-r--r-- 1 spacetimedb spacetimedb 8472211 Nov  8 21:00 00000000000000000000.stdb.log
-rw-r--r-- 1 spacetimedb spacetimedb 4194304 Nov  8 21:00 00000000000000000000.stdb.ofs
-rw-r--r-- 1 spacetimedb spacetimedb       0 Nov  8 20:34 00000000000000000087.stdb.log
-rw-r--r-- 1 spacetimedb spacetimedb       0 Nov  8 20:43 00000000000000000197.stdb.log
-rw-r--r-- 1 spacetimedb spacetimedb       0 Nov  8 20:48 00000000000000000202.stdb.log

Removing those empty files makes the error change to:

Error: failed to get or launch module host: failed to init replica 2000001 for <identity>: out-of-order offset: expected=88 actual=87