Snapshot will lead memory leaks?

General Help/Support
1

do snapshot for my prometheus;

curl -XPOST http://xx.xx.xx.xx:9090/api/v1/admin/tsdb/snapshot

then the prometheus_tsdb_head_chunks increaing and continue , event stop command curl

image
image2656×596 85.7 KB 

Saved profile in /root/pprof/pprof.prometheus.alloc_objects.alloc_space.inuse_objects.inuse_space.005.pb.gz
File: prometheus
Type: inuse_space
Time: Dec 3, 2021 at 1:37pm (CST)
Entering interactive mode (type "help" for commands, "o" for options)
(pprof) top
Showing nodes accounting for 72.23GB, 78.38% of 92.15GB total
Dropped 468 nodes (cum <= 0.46GB)
Showing top 10 nodes out of 71
      flat  flat%   sum%        cum   cum%
   26.65GB 28.92% 28.92%    26.66GB 28.93%  github.com/prometheus/prometheus/tsdb.(*memSeries).mmapCurrentHeadChunk
    7.39GB  8.02% 36.94%    14.52GB 15.75%  github.com/prometheus/prometheus/tsdb/record.(*Decoder).Series
    7.27GB  7.89% 44.83%     7.27GB  7.89%  github.com/prometheus/prometheus/tsdb/encoding.(*Decbuf).UvarintStr (inline)
    7.07GB  7.68% 52.51%     7.07GB  7.68%  github.com/prometheus/prometheus/pkg/labels.(*Builder).Labels
    5.61GB  6.09% 58.60%     5.64GB  6.12%  github.com/prometheus/prometheus/pkg/textparse.(*PromParser).Metric
    4.75GB  5.15% 63.75%     4.75GB  5.15%  github.com/prometheus/prometheus/scrape.newScrapePool.func1
    4.08GB  4.42% 68.17%     5.08GB  5.51%  github.com/prometheus/prometheus/tsdb.newMemSeries
    3.61GB  3.91% 72.09%     3.61GB  3.91%  github.com/prometheus/prometheus/tsdb.(*txRing).add
    2.92GB  3.16% 75.25%     2.92GB  3.16%  github.com/prometheus/prometheus/tsdb/index.(*MemPostings).addFor
    2.88GB  3.12% 78.38%     2.88GB  3.12%  github.com/prometheus/prometheus/tsdb/chunkenc.NewXORChunk
2

more information in snapshot will lead memory leaks? · Issue #9931 · prometheus/prometheus · GitHub