mirror of https://github.com/prometheus/prometheus
3410559c1b
Let older head blocks be compacted once the newest once has samples at 50% of its total range. This allows the memory of the compacted blocks to be released and garbage collected before a new head block gets created. Thereby the number of head blocks is 1 or 2 instead of 2 or 3 and memory spikes are reduced. |
||
---|---|---|
Documentation/format | ||
chunks | ||
cmd/tsdb | ||
labels | ||
test | ||
testdata | ||
tsdbutil | ||
.travis.yml | ||
LICENSE | ||
README.md | ||
block.go | ||
block_test.go | ||
chunks.go | ||
chunks_test.go | ||
compact.go | ||
db.go | ||
db_test.go | ||
db_unix.go | ||
db_windows.go | ||
encoding_helpers.go | ||
head.go | ||
head_test.go | ||
index.go | ||
index_test.go | ||
postings.go | ||
postings_test.go | ||
querier.go | ||
querier_test.go | ||
tombstones.go | ||
tombstones_test.go | ||
wal.go | ||
wal_test.go |
README.md
TSDB
This repository contains the new Prometheus storage layer that will be used in its 2.0 release.
A writeup of its design can be found here.