lib/db: Use SipHash to deal with hash collision in GC (#6826)

If the GC finds a key k that it wants to keep, it records that in a
Bloom filter. If a key k' can be removed but its hash collides with k,
it will be kept. Since the old Bloom filter code was completely
deterministic, the next run would encounter the same collision, assuming
k must still be kept.

A randomized hash function that uses all the SHA-256 bits solves this
problem: the second run has a non-zero probability of removing k', as
long as the Bloom filter is not completely full.
This commit is contained in:
greatroar
2020-07-11 09:36:09 +02:00
committed by GitHub
parent bbda58a29f
commit 9f92f8c609
3 changed files with 40 additions and 16 deletions
+2
View File
@@ -70,6 +70,8 @@ github.com/d4l3k/messagediff v1.2.1/go.mod h1:Oozbb1TVXFac9FtSIxHBMnBCq2qeH/2KkE
github.com/davecgh/go-spew v1.1.0/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/davecgh/go-spew v1.1.1 h1:vj9j/u1bqnvCEfJOwUhtlOARqs3+rkHYY13jYWTU97c=
github.com/davecgh/go-spew v1.1.1/go.mod h1:J7Y8YcW2NihsgmVo/mv3lAwl/skON4iLHjSsI+c5H38=
github.com/dchest/siphash v1.2.1 h1:4cLinnzVJDKxTCl9B01807Yiy+W7ZzVHj/KIroQRvT4=
github.com/dchest/siphash v1.2.1/go.mod h1:q+IRvb2gOSrUnYoPqHiyHXS0FOBBOdl6tONBlVnOnt4=
github.com/dgraph-io/badger v1.6.1 h1:w9pSFNSdq/JPM1N12Fz/F/bzo993Is1W+Q7HjPzi7yg=
github.com/dgraph-io/badger/v2 v2.0.3 h1:inzdf6VF/NZ+tJ8RwwYMjJMvsOALTHYdozn0qSl6XJI=
github.com/dgraph-io/badger/v2 v2.0.3/go.mod h1:3KY8+bsP8wI0OEnQJAKpd4wIJW/Mm32yw2j/9FUVnIM=