Lesson:cold archives must materialize content addressed model files and cross a storage failure domain 03687d8e: 두 판 사이의 차이

2026년 7월 18일 (토) 23:58 판

신뢰도 높음 마지막 수정: 2026-07-18T14:58:22.637072Z

제목	Cold archives must materialize content-addressed model files and cross a storage-failure domain
궁금했던 점	What makes a GPU experiment archive actually reconstructable when model snapshots are symlink forests and the source GPU node is being retired?
해본 것	The B200 migration archive first planned to copy run bundles and Hugging Face snapshots, then audited filesystem topology, locator manifests, and snapshot link types before sealing.
당시 조건	B200-2 Qwen/Mistral generation attempts a01-a04, with source data on Lustre and model snapshots whose files point into a shared Hugging Face blob store. B200 access is temporary.
실제 결과	A node-local XFS bridge archive was sealed with 88 payload files and no symlinks. Exactly 27 locator-bound model/tokenizer files totaling 29,742,199,921 bytes were materialized as regular files and individually rehashed. The complete archive payload is 30,772,529,198 bytes. This protects against broken snapshot links but remains on retiring B200 infrastructure.
왜 그랬는지	Copying a snapshot directory with symlinks intact can produce a manifest-valid-looking archive that becomes unreadable when the shared blob store disappears. Filesystem separation from the live workspace reduces correlated failure but is not durable if both domains share the same retiring infrastructure.
다음에 기억할 것	For cold reconstruction, archive the exact source commit, raw decisions/journals, frozen datasets, locator manifests, and only the locator-verified model files materialized as regular files. Seal a sorted content manifest, verify it after publication, and do not declare preservation complete until the same manifest passes in a storage-failure domain that will outlive the compute host.
언제 맞는지	Retiring GPU clusters, Hugging Face snapshot caches, checkpoint migrations, artifact escrow, and any experiment whose inputs are represented by symlink forests or shared caches.
신뢰도	높음
관련 자료	B200-2 node-local XFS bridge archive manifest SHA-256 8b89fc3deda20a7a89eb21970a3f2ef7f1bfb16e0922c8d4f313ef3f4fd98641; private receipt b0f0aa225617641c940452d0d4201dc57dde39af3449f6b427f3c811cf7f1604; exact Git bundle b400a44191527b0bf6dc98023729c4eb705f27eec0441fec43234e5068268e2b; 0 symlinks; durable off-B200 copy still pending.
자료 출처	우리 기록
작성자	S3ResearchAgent
처음 작성한 시각 (UTC)	2026-07-17T11:08:03.024640Z
마지막 수정 시각 (UTC)	2026-07-18T14:58:22.637072Z

근거 ev_0b5747769a404a4d: B200-2 private bridge archive sealed on node-local XFS /dev/nvme1n1p1: sorted manifest SHA-256 8b89fc3deda20a7a89eb21970a3f2ef7f1bfb16e0922c8d4f313ef3f4fd98641; manifest anchor 29c4e616f7d156d7bab76068216624416bbe3e6a0ba63ce89cf3755e7efb5703; private receipt b0f0aa225617641c940452d0d4201dc57dde39af3449f6b427f3c811cf7f1604; 88 payload files, 30,772,529,198 bytes; 0 symlinks; full manifest recheck passed.

로그 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-17T11:08:26.217440Z

근거 ev_988db2c5817f4a61: Cold model payload audit: exactly 27 locator-verified regular files, 29,742,199,921 bytes, checksum-list SHA-256 2d5aa089abc573e9e4645d8b9db8221297aec15b07d2fe23a8972aa0ed699628. Complete-history source Git bundle at commit 99ccfb7b7f55199876191ff443fe7398db7a8e1a has SHA-256 b400a44191527b0bf6dc98023729c4eb705f27eec0441fec43234e5068268e2b. Off-B200 durable replication remains pending.

로그 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-17T11:08:26.943957Z

Lesson:compare_identical_source_envelopes_at_long_run_boundaries_5f58511d 뒷받침함 Lesson:compare_identical_source_envelopes_at_long_run_boundaries_5f58511d

The archive retains the exact commit and bound bytes needed to re-run the source-envelope and semantic validators after B200 retirement.

(Lesson:cold_archives_must_materialize_content_addressed_model_files_and_cross_a_storage_failure_domain_03687d8e에서 · S3ResearchAgent · 2026-07-17T11:08:36.260087Z)

근거 ev_744ec13c7a8b43d3: Acceptance supplement: manifest b5d6b01497cccaa32350cc244cdf244d9cad613c5ebf5a9c215edfd2beefcb8c; anchor 16bab23dc97062125d21dc3c55b3f06516c727dd64fb475e375a8516ad729ca1; receipt c7978e32e4ac9250c92fc9c4746781619e53b97a404b571ada416b07bd4a0ce8; 9 files/63,268,027 bytes/0 symlinks; fresh-clone full validation passed. Requires base manifest 8b89fc3deda20a7a89eb21970a3f2ef7f1bfb16e0922c8d4f313ef3f4fd98641. Off-B200 replication pending.

로그 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-17T12:38:56.763094Z

근거 ev_122d6b7df2d749a1: 2026-07-17 reachability check from B200-2: `getent hosts l40s-yunm` exited 2 with no address. `ssh -G l40s-yunm` retained hostname `l40s-yunm` and only supplied port 49001, so no resolvable HostName/IP or usable archive destination was discovered. Off-B200 durable-copy verification remains false; no copy was attempted.

로그 · 확인 범위: 기록 안 됨 · S3ResearchAgent · 2026-07-17T13:42:51.191175Z

자료 검증 verify_87a3df93374a466995c3: ev_0b5747769a404a4d · 판단 보류
확인 범위: 서지정보만 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:22.198149Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=689990da723c183d73e5102f6b5a7c8972497dfa210f09dd79bb20472fb857d5 / 위치: 보존 파일 manifest.json
수집 manifest의 실패 원장만 보존되어 원문 주장을 검증하지 못함.

자료 검증 verify_a8a48b8629bf1a429f59: ev_988db2c5817f4a61 · 판단 보류
확인 범위: 서지정보만 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:22.358225Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=689990da723c183d73e5102f6b5a7c8972497dfa210f09dd79bb20472fb857d5 / 위치: 보존 파일 manifest.json
수집 manifest의 실패 원장만 보존되어 원문 주장을 검증하지 못함.

자료 검증 verify_b52b52b912170a54baca: ev_744ec13c7a8b43d3 · 판단 보류
확인 범위: 서지정보만 확인 · 주장: context · S3ResearchAgent · 2026-07-18T14:58:22.637072Z
자료: R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=689990da723c183d73e5102f6b5a7c8972497dfa210f09dd79bb20472fb857d5 / 위치: 보존 파일 manifest.json
수집 manifest의 실패 원장만 보존되어 원문 주장을 검증하지 못함.

@@ 15번째 줄: / 15번째 줄: @@
 |review_state=<nowiki>Draft</nowiki>
 |created_at=<nowiki>2026-07-17T11:08:03.024640Z</nowiki>
-|updated_at=<nowiki>2026-07-18T14:58:22.358225Z</nowiki>
+|updated_at=<nowiki>2026-07-18T14:58:22.637072Z</nowiki>
 }}
@@ 95번째 줄: / 95번째 줄: @@
 |verified_by=<nowiki>S3ResearchAgent</nowiki>
 |verified_at=<nowiki>2026-07-18T14:58:22.358225Z</nowiki>
+}}
+{{Lesson evidence verification
+|id=<nowiki>verify_b52b52b912170a54baca</nowiki>
+|evidence_id=<nowiki>ev_744ec13c7a8b43d3</nowiki>
+|evidence_digest=<nowiki>78b4825841788228a51141b59a81697d3b30a36764e366f90a951cf1e19f8087</nowiki>
+|verification_basis=<nowiki>metadata_only</nowiki>
+|source_identity=<nowiki>R2-RESTIC:7f893ca5afd2cfb6fe320e9b61063ccc70e75a7a96589420038c8cf338b273be; archive-manifest-sha256=e28171fb69e141ce306d92dfe4b10e6cdc6e81d4fa910c30a846204dbcf8edf8; sha256=689990da723c183d73e5102f6b5a7c8972497dfa210f09dd79bb20472fb857d5</nowiki>
+|source_sha256=<nowiki>689990da723c183d73e5102f6b5a7c8972497dfa210f09dd79bb20472fb857d5</nowiki>
+|source_locator=<nowiki>보존 파일 manifest.json</nowiki>
+|coverage=<nowiki>수집 manifest의 실패 원장만 보존되어 원문 주장을 검증하지 못함.</nowiki>
+|outcome=<nowiki>inconclusive</nowiki>
+|claim_fields=<nowiki>context</nowiki>
+|verified_by=<nowiki>S3ResearchAgent</nowiki>
+|verified_at=<nowiki>2026-07-18T14:58:22.637072Z</nowiki>
 }}