hashicorp / hashicorp/packer-plugin-openstack

Feature Request: Add keep_volume option to preserve boot volume artifacts in OpenStack builder

Offen
#151 1 Kommentar 1 Reaktion 0 zugewiesene Personen Auf GitHub ansehen
enhancement
Vorherrschende Sprache
Go
Sterne
29
Forks
26
PR-Merge-Kennzahlen
Keine gemergten PRs in 30 T.

Beschreibung

#### Community Note

Please vote on this issue by adding a 👍 [reaction](https://blog.github.com/2016-03-10-add-reactions-to-pull-requests-issues-and-comments/) to the original issue to help the community and maintainers prioritize this request.
Please do not leave "+1" or "me too" comments, they generate extra noise for issue followers and do not help prioritize the request.
If you are interested in working on this issue or have submitted a pull request, please leave a comment.

#### Description

When use_blockstorage_volume = true, Packer always deletes the created volume during cleanup. There is no way to retain it.

I use Platform9 Managed OpenStack, which supports creating VMs directly from volumes, not just from images. This is a common pattern in environments where volumes are preferred over images for provisioning, because it enables storage locality, faster cloning, and immutable golden volume workflows.

The problem is that Packer does all the hard work, it creates a volume, boots a VM from it, and runs provisioners, but then DELETES the VOLUME at the end. The only artifact it can produce today is a Glance image. If you want a retained volume as your build output, Packer gives you no option.

The workaround right now is to use Terraform or Ansible after a Packer run to re-create the volume from the image. That is an unnecessary extra step. Packer already creates the volume internally ( because we use use_blockstorage_volume =true), but it just always destroys as part of its cleanup process.

I would like to request an optional keep_volume flag (default false) so the volume Packer creates is preserved after a successful build. Existing behavior stays completely unchanged when the flag is not set.

Current behavior (use_blockstorage_volume = true, skip_create_image = false):
get source image --> create volume --> provision VM --> cleanup deletes volume --> Glance image created

Requested behavior (use_blockstorage_volume = true, skip_create_image = true, **keep_volume = true**):
get source image --> create volume --> provision VM --> **cleanup skips volume deletion** --> volume retained (no image created)

Requested behavior (use_blockstorage_volume = true, **keep_volume = true**, skip_create_image = false):
get source image --> create volume --> provision VM --> **cleanup skips volume deletion** --> volume retained + Glance image created

I have implemented this in a fork and it works correctly. Happy to open a PR.

#### Use Case(s)

Use Case(s)
1. Golden volume workflows : Build and harden a boot volume once. Keep it as a reusable template source for VM creation.
2. VM provisioning from volume clones: Clone VMs from preserved golden volumes instead of re-importing images. Keeps storage locality within the same volume type.
3. Immutable artifact lifecycle : Produce versioned, retained build artifacts. Avoid mutating existing template volumes in place.
4. Operational safety: Prevent accidental loss of the build output that today's default cleanup causes. Retain the volume for inspection and controlled rollout.

#### Potential configuration
```
packer {
required_plugins {
openstack = {
source = "github.com/hashicorp/openstack"
version = ">= 1.1.3"
}
}
}

source "openstack" "golden_volume" {
source_image = ""
flavor = ""
image_name = "golden-linux-2026-04-13"

use_blockstorage_volume = true
volume_size = 100
volume_name = "golden-linux-volume-2026-04-13"
volume_type = ""

skip_create_image = true # skip Glance image, volume is the target artifact
keep_volume = true # retain created volume after build
}
```
keep_volume = false (default): existing cleanup behavior, no breaking change
keep_volume = true: created volume is retained, compute instance is still deleted

#### Potential References
OpenStack Cinder docs: https://docs.openstack.org/cinder/latest/
Packer plugin docs: https://developer.hashicorp.com/packer/docs/plugins
OpenStack plugin repo: https://github.com/hashicorp/packer-plugin-openstack

Beitragsleitfaden

Beitragsleitfaden öffnen

Rechercherichtung

Beginne damit, die Handhabung von use_blockstorage_volume durch den OpenStack builder und den Cleanup-Pfad nachzuverfolgen, der das erstellte Volume löscht. Prüfe die Dokumentation des Packer plugin und die angeforderten keep_volume-Fälle. Verifiziere anschließend, dass der standardmäßige Cleanup unverändert bleibt, beibehaltene Volumes erfolgreiche Builds überdauern und Compute-Instanzen weiterhin gelöscht werden.

Vom Indexierungsmodell aus dem Issue-Text verfasst.

Bewertung

Tech-Stack
go
Bereich
cloud, infrastructure
Issue-Typ
Feature
Schwierigkeit
4/5
Geschätzter Aufwand
3-5 Tage
Aktivitätsstatus
Ruhig
Klarheit
Größtenteils klar
Anfängerfreundlichkeit
52/100

Neue Issues direkt in Ihr Postfach

Eine kurze Übersicht über anfängerfreundliche GitHub-Issues.