Scott Lamb b677964d1a properly account for bytes to add with next flush ... This was considering them as 0, so it would under-delete until the next flush them delete all at once. That effectively doubled the number of bytes not yet deleted as they're first transferred to garbage, flushed again, then unlinked.		2018-03-01 13:50:59 -08:00
..
testdata	initial split of database to a separate crate	2018-02-20 23:15:39 -08:00
upgrade	refine 1->3 upgrade process	2018-03-01 09:47:56 -08:00
Cargo.toml	initial split of database to a separate crate	2018-02-20 23:15:39 -08:00
coding.rs	initial split of database to a separate crate	2018-02-20 23:15:39 -08:00
db.rs	properly account for bytes to add with next flush	2018-03-01 13:50:59 -08:00
dir.rs	properly account for bytes to add with next flush	2018-03-01 13:50:59 -08:00
lib.rs	move db upgrade logic into db crate	2018-02-28 21:21:47 -08:00
raw.rs	move open_id from recording_playback to recording	2018-02-28 20:52:43 -08:00
recording.rs	initial split of database to a separate crate	2018-02-20 23:15:39 -08:00
schema.proto	more safety around adding/deleting dirs	2018-03-01 12:24:32 -08:00
schema.rs	initial split of database to a separate crate	2018-02-20 23:15:39 -08:00
schema.sql	move open_id from recording_playback to recording	2018-02-28 20:52:43 -08:00
testutil.rs	properly account for bytes to add with next flush	2018-03-01 13:50:59 -08:00