octoleo
/
restic
mirror of https://github.com/octoleo/restic.git
2
2
Fork 0
Code Issues Releases Wiki Activity

Only repack small files if there are multiple of them

This commit is contained in:
Michael Eischer 2022-07-03 00:16:04 +02:00
parent 1b076cda97
commit 324935cb80
1 changed files with 18 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

22
cmd/restic/cmd_prune.go
View File

@ -71,7 +71,7 @@ func addPruneOptions(c *cobra.Command) {
f.StringVar(&pruneOptions.MaxUnused, "max-unused", "5%", "tolerate given `limit` of unused data (absolute value in bytes with suffixes k/K, m/M, g/G, t/T, a value in % or the word 'unlimited')")
f.StringVar(&pruneOptions.MaxRepackSize, "max-repack-size", "", "maximum `size` to repack (allowed suffixes: k/K, m/M, g/G, t/T)")
f.BoolVar(&pruneOptions.RepackCachableOnly, "repack-cacheable-only", false, "only repack packs which are cacheable")
f.BoolVar(&pruneOptions.RepackSmall, "repack-small", false, "also repack small packs")
f.BoolVar(&pruneOptions.RepackSmall, "repack-small", false, "repack too small packs")
f.BoolVar(&pruneOptions.RepackUncompressed, "repack-uncompressed", false, "repack all uncompressed data")
}
@ -424,8 +424,10 @@ func decidePackAction(ctx context.Context, opts PruneOptions, gopts GlobalOption
repackPacks := restic.NewIDSet()
var repackCandidates []packInfoWithID
var repackSmallCandidates []packInfoWithID
repoVersion := repo.Config().Version
targetPackSize := repo.PackSize()
// consider files with at least 80% of the target size as large enough
targetPackSize := repo.PackSize() / 5 * 4
// loop over all packs and decide what to do
bar := newProgressMax(!gopts.Quiet, uint64(len(indexPack)), "packs processed")
@ -482,8 +484,12 @@ func decidePackAction(ctx context.Context, opts PruneOptions, gopts GlobalOption
stats.packs.keep++
case p.unusedBlobs == 0 && p.tpe != restic.InvalidBlob && !mustCompress && packIsLargeEnough:
// All blobs in pack are used and not mixed => keep pack!
stats.packs.keep++
if packIsLargeEnough {
// All blobs in pack are used and not mixed => keep pack!
stats.packs.keep++
} else {
repackSmallCandidates = append(repackSmallCandidates, packInfoWithID{ID: id, packInfo: p})
}
default:
// all other packs are candidates for repacking
@ -526,6 +532,14 @@ func decidePackAction(ctx context.Context, opts PruneOptions, gopts GlobalOption
}
}
if len(repackSmallCandidates) < 10 {
// too few small files to be worth the trouble, this also prevents endlessly repacking
// if there is just a single pack file below the target size
stats.packs.keep += uint(len(repackSmallCandidates))
} else {
repackCandidates = append(repackCandidates, repackSmallCandidates...)
}
// Sort repackCandidates such that packs with highest ratio unused/used space are picked first.
// This is equivalent to sorting by unused / total space.
// Instead of unused[i] / used[i] > unused[j] / used[j] we use