TrueCloudLab/distribution
13
0
Fork 3
Code Issues 2 Pull requests Projects Releases 2 Packages Wiki Activity Actions

Update the gc documentation.

Browse source 
Signed-off-by: Richard Scothern <richard.scothern@gmail.com>
This commit is contained in:
Richard Scothern 2016-03-29 10:47:22 -07:00
parent 4324b70c50
commit eb0b7f0173
2 changed files with 27 additions and 15 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

35
docs/gc.md
View file

@ -8,21 +8,34 @@ keywords = ["registry, garbage, images, tags, repository, distribution"]
# What Garbage Collection Does
Garbage collection is a process that delete blobs to which no manifests refer.
It runs in two phases. First, in the 'mark' phase, the process scans all the
manifests in the registry. From these manifests, it constructs a set of content
address digests. This set is the 'mark set' and denotes the set of blobs to *not*
delete. Secondly, in the 'sweep' phase, the process scans all the blobs and if
a blob's content address digest is not in the mark set, the process will delete
it.
"Garbage collection deletes blobs which no manifests reference. Manifests and
blobs which are deleted by their digest through the Registry API will become
eligible for garbage collection, but the actual blobs will not be removed from
storage until garbage collection is run.
# How Garbage Collection Works
Garbage collection runs in two phases. First, in the 'mark' phase, the process
scans all the manifests in the registry. From these manifests, it constructs a
set of content address digests. This set is the 'mark set' and denotes the set
of blobs to *not* delete. Secondly, in the 'sweep' phase, the process scans all
the blobs and if a blob's content address digest is not in the mark set, the
process will delete it.
> **NOTE** You should ensure that the registry is in read-only mode or not running at
> all. If you were to upload an image while garbage collection is running, there is the
> risk that the image's layers will be mistakenly deleted, leading to a corrupted image.
This type of garbage collection is known as stop-the-world garbage collection. In
future registry versions the intention is that garbage collection will be an
automated background action and this manual process will no longer apply.
# How to Run
You can run garbage collection by running
docker run --rm registry-image-name garbage-collect /etc/docker/registry/config.yml
`docker run --rm registry-image-name garbage-collect /etc/docker/registry/config.yml`
Additionally, garbage collection can be run in `dry-run` mode, which will print
the progress of the mark and sweep phases without removing any data.
NOTE: You should ensure that the registry itself is in read-only mode or not running at
all. If you were to upload an image while garbage collection is running, there is the
risk that the image's layers will be mistakenly deleted, leading to a corrupted image.

7
registry/garbagecollect.go
View file

@ -19,8 +19,7 @@ import (
func emit(format string, a ...interface{}) {
if dryRun {
fmt.Printf(format, a...)
fmt.Println("")
fmt.Printf(format+"\n", a...)
}
}
@ -122,8 +121,8 @@ func markAndSweep(ctx context.Context, storageDriver driver.StorageDriver, regis
// Construct vacuum
vacuum := storage.NewVacuum(ctx, storageDriver)
for dgst := range deleteSet {
emit("blob eligible for deletion: %s", dgst)
if dryRun {
emit("deleting %s", dgst)
continue
}
err = vacuum.RemoveBlob(string(dgst))
@ -169,7 +168,7 @@ var GCCmd = &cobra.Command{
k, err := libtrust.GenerateECP256PrivateKey()
if err != nil {
fmt.Fprintf(os.Stderr, "%s", err)
fmt.Fprint(os.Stderr, err)
os.Exit(1)
}