v0.1.0

What's Changed

• feat: scaffold monorepo layout by @Ronkahn21 in #1
• build: switch license headers to SPDX format and enforce in CI by @Ronkahn21 in #5
• docs: add Apache-2.0 LICENSE and DCO CONTRIBUTING.md by @danbar2 in #6
• feat: migrate snapshot code from Dynamo by @Ronkahn21 in #7
• fix: move NVIDIA copyright to top of LICENSE, restore APPENDIX placeholder by @Ronkahn21 in #8
• ci: add CodeRabbit configuration by @Ronkahn21 in #10
• chore: mark dead code synced from Dynamo with TODO comments by @Ronkahn21 in #11
• docs: add project README by @Ronkahn21 in #3
• Create SECURITY.md by @nvda-mesharma in #15
• Update readme by @itayvallach in #14
• fix: re-sync restore, overlay, and sidecar fixes from Dynamo by @Ronkahn21 in #17
• ci: add release pipeline for Docker images and Helm chart by @Ronkahn21 in #19
• build(deps): bump actions/checkout from 4 to 7 by @dependabot[bot] in #20
• build: fix agent cross-compilation and Helm operator config by @Ronkahn21 in #9
• feat(operator): enable manager leader election by @Ronkahn21 in #28
• style: gofmt operator manager main.go by @Ronkahn21 in #30
• chore: add CODEOWNERS by @Ronkahn21 in #33
• fix(rbac): add events create+patch for operator and agent by @Ronkahn21 in #34
• build(deps): bump azure/setup-helm from 4.3.1 to 5.0.1 by @dependabot[bot] in #24
• build(deps): bump docker/login-action from 3.7.0 to 4.5.1 by @dependabot[bot] in #23
• feat(operator): upgrade CRDs from an init container by @danbar2 in #32
• chore: add Hutm to CODEOWNERS by @Ronkahn21 in #35
• build(deps): pin deps to clear corporate CVE scan by @Ronkahn21 in #38
• build(operator): use NVIDIA-managed distroless base image by @danbar2 in #39
• fix(agent): enter PID namespace for GPU UUID nvidia-smi probe by @Ronkahn21 in #42
• chore(agent): add //go:build linux constraints to Linux-only files by @oleg-kushniriov in #40
• build: ship corresponding source and attribution inside both images by @danbar2 in #47
• feat: migrate target-containers to PodReference.Containers by @oleg-kushniriov in #50
• feat(agent): add ns-bind-mount C helper and Go mount package by @oleg-kushniriov in #51
• feat: Sync/dynamo 12936 by @oleg-kushniriov in #52
• feat(agent): restore ns mount by @oleg-kushniriov in #53
• test: add Snapshot e2e infra skeleton by @shayasoolin in #41
• ci: allow arm64 e2e runner tooling by @shayasoolin in #54
• fix(ci): allow release artifact publishing by @Ronkahn21 in #55
• test: update e2e PodSnapshot spec by @shayasoolin in #56
• build(deps): bump docker/login-action from 4.5.1 to 4.6.0 by @dependabot[bot] in #43
• build(deps): bump actions/setup-go from 5 to 7 by @dependabot[bot] in #21
• build(deps): bump docker/setup-buildx-action from 3.12.0 to 4.2.0 by @dependabot[bot] in #22
• fix: preserve CUDA launch-job identity across restore by @oleg-kushniriov in #57
• fix(agent): remap clone-conflicting sockets during restore by @oleg-kushniriov in #58
• Sync/dynamo pr 13252 by @oleg-kushniriov in #59
• build(agent): bump Go 1.26.5 references to 1.26.6 by @oleg-kushniriov in #60
• refactor: rename control-dir env to SNAPSHOT_CONTROL_DIR by @oleg-kushniriov in #64
• feat: add SnapshotJob v1alpha1 types and CRD by @oleg-kushniriov in #65
• perf(agent): increase rootfs extraction tar blocking factor by @galletas1712 in #71
• fix(agent): clear stale restore sentinel before restore by @galletas1712 in #72
• chore: add hhzhang16 to CODEOWNERS by @galletas1712 in #76
• fix(agent): stage rootfs-diff.tar locally before extract by @galletas1712 in #73
• fix(agent): keep private CRIU restore images on local disk by @galletas1712 in #74
• fix(agent): name dump and restore timing phases after the work by @galletas1712 in #75
• feat(agent): own checkpoint PVC and mount restore artifacts by @Ronkahn21 in #61
• Snapshotjob source job by @oleg-kushniriov in #68
• fix(agent): release checkpoint after ready status by @galletas1712 in #77
• create the SnapshotJob PodSnapshot and derive Captured by @oleg-kushniriov in #70
• feat: wire the SnapshotJob completion gate, failure handling, and cleanup by @oleg-kushniriov in #80
• Use PodSnapshot references for restore by @Ronkahn21 in #66
• feat!: remove leave-running and complete SnapshotJob from the capture result by @oleg-kushniriov in #101
• refactor(operator): one signal owner per SnapshotJob condition by @oleg-kushniriov in #106
• refactor(operator): watch all pods with a field-stripped cache by @oleg-kushniriov in #123
• build(deps): bump docker/setup-buildx-action from 4.2.0 to 4.3.0 by @dependabot[bot] in #105
• Add parallel multi-container restore by @Ronkahn21 in #113
• fix(operator): wait for CRD establishment in crd-installer by @oleg-kushniriov in #117
• test(e2e): add SnapshotJob capture and restore coverage by @oleg-kushniriov in #81
• refactor(operator): drop CheckpointJob naming from the snapshot protocol by @oleg-kushniriov in #134
• docs(readme): add SnapshotJob to APIs table by @oleg-kushniriov in #139
• feat: clean up PodSnapshotContent artifacts by @Ronkahn21 in #138
• feat(api): publish restore outcome classifier by @julienmancuso in #130
• feat(api): propagate SnapshotJob PodSnapshot metadata by @julienmancuso in #129
• feat(api): publish restore Pod contract by @julienmancuso in #131
• ci: support versioned release branches by @Ronkahn21 in #150
• fix(chart): backport OpenShift SCC admission support by @Ronkahn21 in #158
• docs: Snapshot alpha documentation (backport #132) by @Ronkahn21 in #183
• ci(e2e): pin auto-resolved images to checked-out commit (backport #154) by @Ronkahn21 in #184
• docs: backport #159, #161, #168, and #169 to release/0.1 by @Ronkahn21 in #190
• docs(chart): clarify OpenShift installation by @Ronkahn21 in #194
• docs(guides): backport #170 (SNAPSHOT_RESTORE_STANDBY) to release/0.1 by @oleg-kushniriov in #197
• 📝 RUN-42620: align the usage guides on one namespace, image flow, and restore placeholder (backport #191) by @leeZardNav in #196
• fix(agent): rewrite bound abstract UNIX sockets on clone restore (backport #167) by @galletas1712 in #200
• 📝 docs(guides): verify TensorRT-LLM without a GPU driver (backport #224) by @leeZardNav in #225
• docs(guides): align framework examples (backport #171) by @Ronkahn21 in #228

New Contributors

• @danbar2 made their first contribution in #6
• @nvda-mesharma made their first contribution in #15
• @itayvallach made their first contribution in #14
• @shayasoolin made their first contribution in #41
• @julienmancus...

v0.1.0-rc2

What's Changed

• feat: scaffold monorepo layout by @Ronkahn21 in #1
• build: switch license headers to SPDX format and enforce in CI by @Ronkahn21 in #5
• docs: add Apache-2.0 LICENSE and DCO CONTRIBUTING.md by @danbar2 in #6
• feat: migrate snapshot code from Dynamo by @Ronkahn21 in #7
• fix: move NVIDIA copyright to top of LICENSE, restore APPENDIX placeholder by @Ronkahn21 in #8
• ci: add CodeRabbit configuration by @Ronkahn21 in #10
• chore: mark dead code synced from Dynamo with TODO comments by @Ronkahn21 in #11
• docs: add project README by @Ronkahn21 in #3
• Create SECURITY.md by @nvda-mesharma in #15
• Update readme by @itayvallach in #14
• fix: re-sync restore, overlay, and sidecar fixes from Dynamo by @Ronkahn21 in #17
• ci: add release pipeline for Docker images and Helm chart by @Ronkahn21 in #19
• build(deps): bump actions/checkout from 4 to 7 by @dependabot[bot] in #20
• build: fix agent cross-compilation and Helm operator config by @Ronkahn21 in #9
• feat(operator): enable manager leader election by @Ronkahn21 in #28
• style: gofmt operator manager main.go by @Ronkahn21 in #30
• chore: add CODEOWNERS by @Ronkahn21 in #33
• fix(rbac): add events create+patch for operator and agent by @Ronkahn21 in #34
• build(deps): bump azure/setup-helm from 4.3.1 to 5.0.1 by @dependabot[bot] in #24
• build(deps): bump docker/login-action from 3.7.0 to 4.5.1 by @dependabot[bot] in #23
• feat(operator): upgrade CRDs from an init container by @danbar2 in #32
• chore: add Hutm to CODEOWNERS by @Ronkahn21 in #35
• build(deps): pin deps to clear corporate CVE scan by @Ronkahn21 in #38
• build(operator): use NVIDIA-managed distroless base image by @danbar2 in #39
• fix(agent): enter PID namespace for GPU UUID nvidia-smi probe by @Ronkahn21 in #42
• chore(agent): add //go:build linux constraints to Linux-only files by @oleg-kushniriov in #40
• build: ship corresponding source and attribution inside both images by @danbar2 in #47
• feat: migrate target-containers to PodReference.Containers by @oleg-kushniriov in #50
• feat(agent): add ns-bind-mount C helper and Go mount package by @oleg-kushniriov in #51
• feat: Sync/dynamo 12936 by @oleg-kushniriov in #52
• feat(agent): restore ns mount by @oleg-kushniriov in #53
• test: add Snapshot e2e infra skeleton by @shayasoolin in #41
• ci: allow arm64 e2e runner tooling by @shayasoolin in #54
• fix(ci): allow release artifact publishing by @Ronkahn21 in #55
• test: update e2e PodSnapshot spec by @shayasoolin in #56
• build(deps): bump docker/login-action from 4.5.1 to 4.6.0 by @dependabot[bot] in #43
• build(deps): bump actions/setup-go from 5 to 7 by @dependabot[bot] in #21
• build(deps): bump docker/setup-buildx-action from 3.12.0 to 4.2.0 by @dependabot[bot] in #22
• fix: preserve CUDA launch-job identity across restore by @oleg-kushniriov in #57
• fix(agent): remap clone-conflicting sockets during restore by @oleg-kushniriov in #58
• Sync/dynamo pr 13252 by @oleg-kushniriov in #59
• build(agent): bump Go 1.26.5 references to 1.26.6 by @oleg-kushniriov in #60
• refactor: rename control-dir env to SNAPSHOT_CONTROL_DIR by @oleg-kushniriov in #64
• feat: add SnapshotJob v1alpha1 types and CRD by @oleg-kushniriov in #65
• perf(agent): increase rootfs extraction tar blocking factor by @galletas1712 in #71
• fix(agent): clear stale restore sentinel before restore by @galletas1712 in #72
• chore: add hhzhang16 to CODEOWNERS by @galletas1712 in #76
• fix(agent): stage rootfs-diff.tar locally before extract by @galletas1712 in #73
• fix(agent): keep private CRIU restore images on local disk by @galletas1712 in #74
• fix(agent): name dump and restore timing phases after the work by @galletas1712 in #75
• feat(agent): own checkpoint PVC and mount restore artifacts by @Ronkahn21 in #61
• Snapshotjob source job by @oleg-kushniriov in #68
• fix(agent): release checkpoint after ready status by @galletas1712 in #77
• create the SnapshotJob PodSnapshot and derive Captured by @oleg-kushniriov in #70
• feat: wire the SnapshotJob completion gate, failure handling, and cleanup by @oleg-kushniriov in #80
• Use PodSnapshot references for restore by @Ronkahn21 in #66
• feat!: remove leave-running and complete SnapshotJob from the capture result by @oleg-kushniriov in #101
• refactor(operator): one signal owner per SnapshotJob condition by @oleg-kushniriov in #106
• refactor(operator): watch all pods with a field-stripped cache by @oleg-kushniriov in #123
• build(deps): bump docker/setup-buildx-action from 4.2.0 to 4.3.0 by @dependabot[bot] in #105
• Add parallel multi-container restore by @Ronkahn21 in #113
• fix(operator): wait for CRD establishment in crd-installer by @oleg-kushniriov in #117
• test(e2e): add SnapshotJob capture and restore coverage by @oleg-kushniriov in #81
• refactor(operator): drop CheckpointJob naming from the snapshot protocol by @oleg-kushniriov in #134
• docs(readme): add SnapshotJob to APIs table by @oleg-kushniriov in #139
• feat: clean up PodSnapshotContent artifacts by @Ronkahn21 in #138
• feat(api): publish restore outcome classifier by @julienmancuso in #130
• feat(api): propagate SnapshotJob PodSnapshot metadata by @julienmancuso in #129
• feat(api): publish restore Pod contract by @julienmancuso in #131
• ci: support versioned release branches by @Ronkahn21 in #150
• fix(chart): backport OpenShift SCC admission support by @Ronkahn21 in #158
• docs: Snapshot alpha documentation (backport #132) by @Ronkahn21 in #183
• ci(e2e): pin auto-resolved images to checked-out commit (backport #154) by @Ronkahn21 in #184
• docs: backport #159, #161, #168, and #169 to release/0.1 by @Ronkahn21 in #190
• docs(chart): clarify OpenShift installation by @Ronkahn21 in #194
• docs(guides): backport #170 (SNAPSHOT_RESTORE_STANDBY) to release/0.1 by @oleg-kushniriov in #197
• 📝 RUN-42620: align the usage guides on one namespace, image flow, and restore placeholder (backport #191) by @leeZardNav in #196
• fix(agent): rewrite bound abstract UNIX sockets on clone restore (backport #167) by @galletas1712 in #200

New Contributors

• @danbar2 made their first contribution in #6
• @nvda-mesharma made their first contribution in #15
• @itayvallach made their first contribution in #14
• @shayasoolin made their first contribution in #41
• @julienmancuso made their first contribution in #130

Full Changelog: https://github.com/ai-dynamo/snapshot/commits/v0.1.0-rc2

Snapshot v0.1.0-rc.1

What's Changed

• feat: scaffold monorepo layout by @Ronkahn21 in #1
• build: switch license headers to SPDX format and enforce in CI by @Ronkahn21 in #5
• docs: add Apache-2.0 LICENSE and DCO CONTRIBUTING.md by @danbar2 in #6
• feat: migrate snapshot code from Dynamo by @Ronkahn21 in #7
• fix: move NVIDIA copyright to top of LICENSE, restore APPENDIX placeholder by @Ronkahn21 in #8
• ci: add CodeRabbit configuration by @Ronkahn21 in #10
• chore: mark dead code synced from Dynamo with TODO comments by @Ronkahn21 in #11
• docs: add project README by @Ronkahn21 in #3
• Create SECURITY.md by @nvda-mesharma in #15
• Update readme by @itayvallach in #14
• fix: re-sync restore, overlay, and sidecar fixes from Dynamo by @Ronkahn21 in #17
• ci: add release pipeline for Docker images and Helm chart by @Ronkahn21 in #19
• build(deps): bump actions/checkout from 4 to 7 by @dependabot[bot] in #20
• build: fix agent cross-compilation and Helm operator config by @Ronkahn21 in #9
• feat(operator): enable manager leader election by @Ronkahn21 in #28
• style: gofmt operator manager main.go by @Ronkahn21 in #30
• chore: add CODEOWNERS by @Ronkahn21 in #33
• fix(rbac): add events create+patch for operator and agent by @Ronkahn21 in #34
• build(deps): bump azure/setup-helm from 4.3.1 to 5.0.1 by @dependabot[bot] in #24
• build(deps): bump docker/login-action from 3.7.0 to 4.5.1 by @dependabot[bot] in #23
• feat(operator): upgrade CRDs from an init container by @danbar2 in #32
• chore: add Hutm to CODEOWNERS by @Ronkahn21 in #35
• build(deps): pin deps to clear corporate CVE scan by @Ronkahn21 in #38
• build(operator): use NVIDIA-managed distroless base image by @danbar2 in #39
• fix(agent): enter PID namespace for GPU UUID nvidia-smi probe by @Ronkahn21 in #42
• chore(agent): add //go:build linux constraints to Linux-only files by @oleg-kushniriov in #40
• build: ship corresponding source and attribution inside both images by @danbar2 in #47
• feat: migrate target-containers to PodReference.Containers by @oleg-kushniriov in #50
• feat(agent): add ns-bind-mount C helper and Go mount package by @oleg-kushniriov in #51
• feat: Sync/dynamo 12936 by @oleg-kushniriov in #52
• feat(agent): restore ns mount by @oleg-kushniriov in #53
• test: add Snapshot e2e infra skeleton by @shayasoolin in #41
• ci: allow arm64 e2e runner tooling by @shayasoolin in #54
• fix(ci): allow release artifact publishing by @Ronkahn21 in #55
• test: update e2e PodSnapshot spec by @shayasoolin in #56
• build(deps): bump docker/login-action from 4.5.1 to 4.6.0 by @dependabot[bot] in #43
• build(deps): bump actions/setup-go from 5 to 7 by @dependabot[bot] in #21
• build(deps): bump docker/setup-buildx-action from 3.12.0 to 4.2.0 by @dependabot[bot] in #22
• fix: preserve CUDA launch-job identity across restore by @oleg-kushniriov in #57
• fix(agent): remap clone-conflicting sockets during restore by @oleg-kushniriov in #58
• Sync/dynamo pr 13252 by @oleg-kushniriov in #59
• build(agent): bump Go 1.26.5 references to 1.26.6 by @oleg-kushniriov in #60
• refactor: rename control-dir env to SNAPSHOT_CONTROL_DIR by @oleg-kushniriov in #64
• feat: add SnapshotJob v1alpha1 types and CRD by @oleg-kushniriov in #65
• perf(agent): increase rootfs extraction tar blocking factor by @galletas1712 in #71
• fix(agent): clear stale restore sentinel before restore by @galletas1712 in #72
• chore: add hhzhang16 to CODEOWNERS by @galletas1712 in #76
• fix(agent): stage rootfs-diff.tar locally before extract by @galletas1712 in #73
• fix(agent): keep private CRIU restore images on local disk by @galletas1712 in #74
• fix(agent): name dump and restore timing phases after the work by @galletas1712 in #75
• feat(agent): own checkpoint PVC and mount restore artifacts by @Ronkahn21 in #61
• Snapshotjob source job by @oleg-kushniriov in #68
• fix(agent): release checkpoint after ready status by @galletas1712 in #77
• create the SnapshotJob PodSnapshot and derive Captured by @oleg-kushniriov in #70
• feat: wire the SnapshotJob completion gate, failure handling, and cleanup by @oleg-kushniriov in #80
• Use PodSnapshot references for restore by @Ronkahn21 in #66
• feat!: remove leave-running and complete SnapshotJob from the capture result by @oleg-kushniriov in #101
• refactor(operator): one signal owner per SnapshotJob condition by @oleg-kushniriov in #106
• refactor(operator): watch all pods with a field-stripped cache by @oleg-kushniriov in #123
• build(deps): bump docker/setup-buildx-action from 4.2.0 to 4.3.0 by @dependabot[bot] in #105
• Add parallel multi-container restore by @Ronkahn21 in #113
• fix(operator): wait for CRD establishment in crd-installer by @oleg-kushniriov in #117
• test(e2e): add SnapshotJob capture and restore coverage by @oleg-kushniriov in #81
• refactor(operator): drop CheckpointJob naming from the snapshot protocol by @oleg-kushniriov in #134
• docs(readme): add SnapshotJob to APIs table by @oleg-kushniriov in #139
• feat: clean up PodSnapshotContent artifacts by @Ronkahn21 in #138
• feat(api): publish restore outcome classifier by @julienmancuso in #130
• feat(api): propagate SnapshotJob PodSnapshot metadata by @julienmancuso in #129
• feat(api): publish restore Pod contract by @julienmancuso in #131
• ci: support versioned release branches by @Ronkahn21 in #150

New Contributors

• @danbar2 made their first contribution in #6
• @nvda-mesharma made their first contribution in #15
• @itayvallach made their first contribution in #14
• @dependabot[bot] made their first contribution in #20
• @oleg-kushniriov made their first contribution in #40
• @shayasoolin made their first contribution in #41
• @galletas1712 made their first contribution in #71
• @julienmancuso made their first contribution in #130

Full Changelog: https://github.com/ai-dynamo/snapshot/commits/v0.1.0-rc.1

v0.1.0-alpha.1

build(agent): bump Go 1.26.5 references to 1.26.6 (#60)

agent/Dockerfile's GO_VERSION build-arg default and hack/tools.mk's
GO_VERSION make variable were still 1.26.5, predating go.mod's
`go 1.26.6` requirement bumped in cd5ca84. hack/tools.mk's GO_VERSION
is the actual value docker-build-agent passes as --build-arg (it
overrides the Dockerfile default), so the container's `go mod download`
failed with "go.mod requires go >= 1.26.6 (running go 1.26.5;
GOTOOLCHAIN=local)".

Signed-off-by: Oleg Kushniriov <okushniriov@nvidia.com>