Glance & Cinder Image-Volume Conversion Prompt
Design boot-from-volume and image↔volume conversion flows across Glance and Cinder — enabling the image-volume cache, format conversion, and fast Ceph clone-backed boots.
- Target user
- Operators optimizing image and boot-from-volume pipelines
- Difficulty
- Intermediate
- Tools
- Claude, ChatGPT
The prompt
You are a senior OpenStack operator who has tuned the Glance/Cinder image and boot-from-volume pipeline so instances boot in seconds instead of waiting on multi-GB image copies. I will provide: - Glance store backend (Ceph RBD, file, Swift) and image formats in use (qcow2, raw) - Cinder backends and whether Cinder + Glance share the same Ceph pool/cluster - `cinder image-volume cache` status and `volume create --image` behavior today - Symptoms: slow boot-from-volume, full conversion on every boot, image-format mismatches - Typical image sizes and instance launch rate Your job: 1. **Map the flows** — clarify the distinct operations people conflate: `openstack image create --volume` (volume→image), `cinder create --image` (image→volume), boot-from-volume, and `nova image-create` snapshots. State what each actually does and where the bytes move. 2. **Format strategy** — the big win: when Glance and Cinder share Ceph, store images as **raw** so Cinder can do a fast RBD COW clone instead of a full qcow2→raw conversion per boot. Explain the storage cost vs boot-speed trade. 3. **Image-volume cache** — enable and size `cinder image-volume cache` so the first volume-from-image seeds a cache entry and subsequent boots clone instantly; cover eviction limits and per-backend behavior. 4. **Ceph clone path** — verify the COW clone path is actually taken (not a fallback full copy); the common misconfig where mismatched pools or non-raw images silently defeat cloning. 5. **Conversion safety** — where conversion runs (cinder-volume host), its disk/CPU/temp-space cost, and how a flood of conversions can fill `image_conversion_dir`. 6. **Boot-from-volume policy** — when to force BFV (persistence, live-migration, large root disks) vs local ephemeral, and flavor/quota implications. 7. **Validation** — measure cold vs warm boot-from-volume time, confirm clones via Ceph (`rbd children`), and watch conversion temp space. Output as: (a) flow map, (b) format + shared-pool recommendation, (c) image-volume cache config, (d) a checklist to confirm COW cloning is active, (e) a boot-time validation plan. Bias toward: raw images on shared Ceph, verified COW cloning, and protecting conversion temp space from floods.
Run this prompt with AI
Test it, get an AI-improved version, or compare models — live in the Prompt Workspace. No copy-paste.
Related prompts
-
Glance Cinder-Store & Image-Volume Cache Tuning Prompt
Tune the Glance cinder store and Cinder's image-volume cache so booting from image skips redundant downloads and uses backend-side cloning.
-
Cinder Backend Capacity Rebalance Runbook Prompt
Plan a controlled rebalance of Cinder volumes across storage backends when one backend is near-full and another is idle — using volume migration and retype — without stalling tenant I/O or letting the scheduler keep filling the hot backend.
-
Cinder Volume Stuck-State Recovery Prompt
Safely diagnose Cinder volumes stuck in transitional states (creating, attaching, detaching, error_deleting, in-use after VM deletion) by correlating cinder-volume logs, backend driver state, and Nova attachment records before any reset-state.
-
Glance Image Upload & Store Failure Triage Prompt
Diagnose Glance image problems — uploads stuck in saving/queued, images that won't boot due to bad properties or format, store backend errors (Ceph/Swift/file), and signature/conversion failures — before re-uploading or deleting image data.
More OpenStack prompts & error guides
Browse every OpenStack prompt and troubleshooting guide in one place.
Reading prompts? Get all 500 in one free PDF
500 battle-tested, copy-paste AI prompts engineered by a senior systems engineer — every one with fill-in placeholders and safety/back-out notes. Drop your email and it's yours.
- 500 prompts: Linux · Kubernetes · Terraform · OpenStack · GitLab · Docker · Monitoring · Incident Response
- Instant PDF download — yours free, forever
- Plus one practical AI-workflow email a week (no spam)
Single opt-in · unsubscribe anytime · no spam.