Skip to content

[v0.1.3][Phase 2] Fix payload creation scaling #41

Closed

Description

@yifanxuaaa

Current user-authorized scaling gate — final 2026-09-05 amendment

The user explicitly chose to accept the frozen performance with a strictly < 1.25 normalized adjacent-tier MEDIAN time-growth / actual-work-growth gate and proceed to the host-owned SQLite migration. This supersedes the earlier <1.10 gate and interrupted1.20 proposal. Describe this as a25% scaling-overhead allowance, not proportional scaling.

All other requirements remain unchanged: prescribed114unique samples/seeds1–3, explicit actualwork and every adjacenttier, per-sample15s/runtime/resource limits, no deliberate slowing of smaller tiers, exact applicable semantics and bounded separateverification, authentic Docker/FUSE/no-data-mount setup for the frozen campaign, independent Workspaceclones/nativefresh outputs, cleanup and evidence-bound completion. Preserve rawreceipts and the historical <=1.00/<1.10 misses, individualseed ratios and variability. All30 frozenCandidate3 mediancomparisons clear <1.25; the earlier payload1.105934087 and cross-file mixed1.157331012 misses now clear only the newly amended gate.

Finish remaining frozen assessment/requiredchecks and truthful report updates, then proceed promptly in this same task/worktree with host SDK/coordinator+Workspace/Store/localSQLite and Linux daemon/FUSE/workload via existing authenticated transport. No newHTTP layer, system cache/Docker memory tuning, or separateagent/task/worktree. Host reusable preparation needs protected consistentSQLite masters/backups, independent writable samples and real cache reuse; benchmark topology/source/resource identities and throughput comparisons must be explicit. ExistingDocker-owned results are not host-Store qualification.

The earlier amendment and historical requirements below remain intact as prior records; this paragraph controls the current threshold and sequence.

Current user-authorized scaling gate — 2026-09-05 amendment

For #38 and all four children #41–#44, each required adjacent-tier normalized median time-growth / actual-work-growth ratio must be strictly < 1.10. This permits bounded scaling overhead; it is not a claim of no worse than proportional growth. The user explicitly authorized this amendment in the side conversation and notified the integration owner.

All other gates remain unchanged: original three seeds, 114 unique final candidate samples plus applicable invalidated controls, explicit work dimensions and whole-curve comparisons, every sample within the 15-second product cutoff and resource bounds, no deliberate slowing of smaller tiers, exact applicable semantics and bounded separate verification, Mac + Docker/Linux with real FUSE where applicable and no Docker data mounts, independent Workspace clones, native fresh output Stores, cleanup and evidence-bound issue completion. Individual seed ratios and variability remain visible. This amendment does not automatically close any family.

Older <=1.00 / no-worse-than-proportional clauses below retain the previous strict-gate plan and historical assessments; only their numerical terminal threshold is superseded. Preserve the old raw results and strict-gate failures, and report amended outcomes separately. Candidate2's payload 100→500 median ratio 1.057185870 and cross-file unique ratio 1.053614071 clear this amended median gate; cross-file unique seed3 around1.134 remains an individual variability flag. Complete family qualification remains pending.

Comparison policy amendment — 2026-09-05

The user's latest instruction permits practical historical comparisons and overrides earlier blanket requirements to recollect a matched baseline. Do not rerun or resurrect the old harness merely to obtain a comparison table. Reuse retained baseline receipts/results when the workload and measurement meaning are sufficiently similar; small documented harness differences are acceptable without a new baseline run.

  • Classify comparisons as matched, historical-approximate, or not-comparable, and state the relevant difference in the table. Different source revisions are expected; preserve their provenance. Historical host/split-owner/APFS results may be used as approximate context where the selected product timer still measures comparable work, but must not be relabeled matched Docker/Linux evidence.
  • Judge similarity from actual operation, bytes/files/history depth, seed schedule, timing boundary, and resource scope. Do not introduce an arbitrary percentage tolerance or an extra certification campaign. If a difference could plausibly explain the claimed gain, qualify the conclusion rather than attributing it entirely to the product change.
  • Materially compacted fixtures (compact-v2/compact-v3 versus original large backgrounds), changed measured operations, and incompatible timing/resource scopes are not direct speedup comparisons. Show their absolute results side by side with the difference; leave speedup blank. Do not pool changed low tiers with old high tiers as an unchanged scaling curve.
  • Collect a new baseline with the current harness only for a specific unresolved decision that historical evidence cannot answer, using the smallest necessary selection. A full baseline campaign is not a prerequisite. Retain the prescribed candidate coverage, correctness checks, runtime limits and scaling criteria; this exception concerns baseline reuse, not permission to mark untested cases passing.
  • Derive tables on demand from compact logs; no benchmark rerun is needed to change presentation. See [v0.1.3] Optimize benchmark infrastructure: Docker/FUSE, reusable setup, three modes, compact logs #45 for field mapping. Preserve sample counts and seeds: N repetitions of one seed are not N different seeds. Verification comes from a separate identity-matched receipt, never inferred from performance PASS.

Scope and observed problem

Qualify payload-create across the existing 1/10/100/500 MiB tiers (case IDs retain the m suffix). Payload-read is a regression control only if the implementation change invalidates its route.

Both Exec and Commit show worse-than-proportional growth. Exec medians are 0.186139 → 1.428129 s (7.67×); Commit medians are 0.275633 → 2.291853 s (8.31×). Total 500 MiB product time is 3.744340 s against a baseline-derived proportional bound of 2.378459 s. Corresponding-seed total growth factors are 5.36×, 7.92× and 14.05× for 5× work.

Inspect existing phase/CPU/I/O/admission receipts for both execution and Commit before selecting a fix. These observations establish the symptom, not a proven root cause. Preserve the live public operation and complete stage/publication/installation/cleanup timing boundaries.

Historical compatibility entry point: benchmark/fs-bench-pro/run-payload-create-read.sh.

Final listed-curve coverage: 1 curve(s) × 4 tiers × 3 seeds = 12 candidate performance samples, before invalidated regression controls; reuse valid evidence rather than repeating every sample during iteration.

Workspace track: shared implementation with #44

Work alongside #44 under one Workspace implementation owner after #45 handoff. Own the single-large-file creation/capture path and complete Commit; coordinate changes in shared layerfs-workspace and Store admission code rather than implementing a separate payload engine.

Already adopted in #40: exact metadata cache, sorted Workspace tree adapters, checked admission and carried batches (127 -> 8191 object ceiling, still <4 MiB), staging and exact-result continuation. Content builders and optional single-new-file capture existed earlier. Missing transfer: owned candidate delivery to admission; current deferred candidate may spill and is later copied into admission batches. No general reusable producer pool exists to enable blindly.

Start by reading #45's payload_create_read smoke/setup record and one selected small current-source payload case. Separate Exec/capture, candidate content/tree, copies/spill/readback, admission and continuation. Inspect the historical 100/500 pair only after setup and the selected route are sound. If a useful #44/shared change resolves the complete curve, qualify it without another patch. Preserve this family's one large file, real FUSE write, metadata and fsync-before-close; small-file close batching is not an equivalent workload.

Infrastructure prerequisite and handoff

Execution starts after #45's benchmark-infrastructure migration is complete and its handoff is ready. The current infrastructure task is 01a06da4-c2ba-7013-92cb-392a348dfdfe (Implement LayerFS Phase 2.1). Do not start product optimization, builds or benchmark collection in parallel with that migration. This issue update defines the successor work; it does not launch a handoff agent.

Read #45 and docs/roadmap/0.1/0.1.3/benchmark-infrastructure-optimization-spec.md at the final handoff revision. Consume its actual migrated entrypoints, compact smoke results, slow-case inventory and unsupported/failure dispositions; do not assume a proposed path is already implemented. A small smoke PASS proves the migrated route works, not that larger tiers or this scaling target pass. Reuse valid smoke/proof evidence; do not rerun it merely to begin a new task. Slow or blocked product cases remain explicit diagnosis input, not a reason to repeat known-slow large runs or label them passing.

#40 is the product-mechanism prerequisite. Use its recorded Approach B implementation/handoff, with the namespace <=2.2-second and 10% CPU-reduction targets still explicitly missed. Do not silently turn that MISS into PASS or restart namespace tuning as a gate for this work. Audit actual adoption of its six mechanisms; the completion summary alone does not establish full caller transfer. The user's sequencing is infrastructure completion, then the #38 optimization handoff.

Parent: #38, under #21. Migrated family ID: payload_create_read. Preserve this child's original acceptance criteria and case ownership.

Historical evidence

The table records retained Phase 1 observations, not results on the #40 candidate. Median ratios use unrounded product-time samples. “Over linear” = (larger median / smaller median) / (larger actual work / smaller actual work) - 1. “Seeds over” is descriptive, not statistical confidence.

Use the source/evidence selection in #38 and the final Phase 1 report. Affected baseline sources are 7948df2d, with corrected CDC delete tier-500 seed 3 at e24a3b34. Preserve exact case/seed overlays, source/environment/input identities and historical failures; exclude #39's persistent suppression inventory.

Retained baseline statistics

Exact curve prefix Tier transition Actual work growth Median seconds Time growth Over linear Seeds over
payload-create (MiB suffix m) 100→500 5.000× 0.475692→3.744340 7.871× 57.4% 3/3

Concrete files, shared methods and cost model

Already-shared helpers (PortableMetadataCache, sorted Workspace tree updates and insert_checked_object_batch) are retained. The first candidate refactor, if measurements show copying matters, is DeferredObjectStore::consume_prevalidated_pages -> admit_checked_objects -> commit_workspace_candidate, in crates/layerfs-layerstack-store/src/{objects.rs,workspace.rs}. Consume selected in-memory vectors instead of bytes.to_vec(). Preserve reachable ordering, identity/conflict checks, carried 8,191-object/<4 MiB limits and transaction outcomes. Spill still copies; this does not eliminate candidate replay or create an overlapping pipeline.

Only if larger replay costs justify it, adapt existing Store-owned slab/writer/admission machinery through a narrow nonempty-Store boundary, with layerfs-workspace/src/changes.rs supplying stable selected output. Do not reuse the initializer's empty-Store-only constructor unchanged or add a new public SDK endpoint/pool. lifecycle.rs::rebase_committed is a separate continuation target, not an admission speedup.

Cost model: for N selected objects, B canonical bytes, D existing-conflict bytes and M prior Store objects, admission is candidate visitation plus O(B + D + N log(M + N + 1)) indexed/byte work plus transaction/physical-I/O costs. Additional admission buffers are bounded; full candidate/spill/Workspace/SQLite space is not just the batch limit. Record generated/selected bytes, copies, spill/readback, batch occupancy/transactions and complete Commit/continuation phases. Moving owned pages removes a copy but normally does not change the asymptotic class.

Family-specific boundary: layerfs-workspace/src/capture.rs already has optional single-new-file capture. Preserve and measure it; changes.rs consumes captured/deferred content. #41 is one large synced file, so multi-small-file close batching is not its optimization substitute. Own Workspace changes jointly with #44.

Planning documents in docs/roadmap/0.1/0.1.3/: phase-2-shared-code-layout.md, phase-2.1-mechanism-adoption-audit.md, and workspace-admission-complexity.md. Current local copies are in /Users/yifanxu/Ephemeral-AI-Lab/layerfs/docs/roadmap/0.1/0.1.3/; these edits are local and not claimed committed/pushed. The key instructions are included here so this issue is self-contained. Audit source was 810bb3a58; recheck the final #45 handoff source and paths before editing.

API reduction counts and algorithm consolidation

The reviewed target is fewer independent implementations, not an immediate large reduction of public SDK methods. Inventory at product source 810bb3a58: 26 production Client methods plus two feature-gated verification helpers. Reinventory the final #45 handoff source before editing.

Surface Initial change Later candidate
Public Rust SDK 0 methods removed; 26 remain, with compatible forwarding wrappers and consistent validation/observation One versioned Commit-variant reduction, 26 -> 25 only if no other API changes occur; migrate callers and preserve full publication/presentation status first
FUSE wire protocol Preserve acknowledged readonly Pin and current compatibility 1 identified obsolete request: PinRead, with no current production sender; pruning requires an explicit compatibility disposition and is not a guaranteed net request-count change
Internal Rust methods No validated net deletion count Count the reviewed implementation diff; a shared helper may remove duplicated bodies without reducing total methods
Algorithm area Current implementations Proposed target
Native scheduling/orchestration 3 1 scheduling implementation, with explicit appropriate sinks
Admission accumulation/orchestration 3 1 shared bounded accumulator; operation-specific publication stays explicit
Native regular-file construction bodies 2 1
Workspace inode-record construction bodies 2 1
Workspace candidate planners 3 Eventually 1, after all fallback domains and sparse/resource behavior transfer

These are not additive method-removal counts, performance promises or mandatory deletion quotas. Compatibility wrappers may remain. Do not create a catch-all endpoint or policy-flag framework merely to reduce enum/method names. Do not remove a route until its real callers and supported states transfer; test-only differential oracles are not extra production routes.

This family's relevant target is shared Workspace inode-record construction and owned/checked admission with #44. Existing singular/batch file edits already share an atomic core; do not replace the batch with a loop or add a payload-specific SDK API. Preserve optional single-new-file capture and its one-large-file fsync behavior. Consuming owned memory pages can remove a copy without changing indexed admission complexity; direct replay elimination requires separate evidence. Planner deletion remains conditional on supported-state transfer, not a task to force into the first optimization.

Detailed local plans: docs/roadmap/0.1/0.1.3/api-algorithm-simplification-audit.md, phase-2-shared-code-layout.md, and workspace-admission-complexity.md in /Users/yifanxu/Ephemeral-AI-Lab/layerfs. These documents are local planning edits, not claimed pushed commits. The counts and scope are included here so the issue is self-contained.

Acceptance

  • Diagnose the shared product cause using existing evidence before implementation; coordinate shared fixes across sibling family issues and reuse [v0.1.3][Phase 2.1] Share construction and Workspace staging; reach 2.2s initialization with lower CPU #40's applicable implementation.
  • Qualify every listed curve over all four existing tiers with the original three seeds and prescribed operations. Record actual work denominators and every adjacent-tier normalized ratio; require median time growth / actual work growth <1.25 under the final user amendment.
  • Report each sample, median, mean, min/max and corresponding-seed ratios. Keep unresolved inconsistent results explicitly inconclusive; do not add unapproved tolerance or discard inconvenient samples.
  • Preserve smaller-tier baseline performance and publish absolute before/after times. Do not slow smaller cases to manufacture passing ratios. Every candidate sample must remain within the 15-second declared product-time cutoff.
  • Preserve byte/namespace/Commit/CAS/CDC semantics, failure propagation, cleanup and applicable resource limits. Each file remains <=500 MiB and aggregate logical files <1 GiB at every workload state.
  • Publish source-bound before/after results for this family's listed curves and the specific regression controls invalidated by the changes, including relevant phase/CPU/I/O/memory evidence. Timing alone does not qualify storage savings or release readiness.
  • Run the bounded selected verification required by the changed route, outside performance timing, using the repaired existing companion and independent oracles. Retain explicit failures/timeouts/omissions and reuse identity-matched valid evidence.

Completed for the frozen Docker-owned profile under the explicit final amendment; detailed distributions, historic comparisons/limits and correctness scope are in the completion comment.

Migrated benchmark contract

  • All fixture/reference preparation, sample copying, SDK/coordinator, Store, daemon and product work run inside Docker/Linux. Host orchestration and compact result collection only. No runtime data bind mounts, named/anonymous volumes, Docker socket, APFS clone or host-side measured SDK/Store. Real FUSE/Exec remains where the family operation requires it; pure imports do not gain a synthetic timed Workspace lifecycle.
  • Use the migrated families/<canonical-family-id>/{mod.rs,setup.sh,perf.sh,verify.sh} interfaces backed by shared infrastructure. setup.sh is optional prewarming; perf acquires compatible preparation automatically. Historical wrappers are forwarding/archival compatibility only as recorded by [v0.1.3] Optimize benchmark infrastructure: Docker/FUSE, reusable setup, three modes, compact logs #45; do not build a parallel runner.
  • Default --perf-fast means one complete selected workload sample, not smaller data or weaker checks. --perf-samples N repeats the same selected seed/input sequentially; sample ordinal is not seed, repetition or history depth. --verification is an explicit separate selection through the single shared verifier. Do not use N to invent new seeds or replace the final prescribed seed schedule.
  • Import measurements reuse prepared source files but always initialize a fresh/absent output Store (fresh-output); reject --setup clone. Post-initialization Workspace cases may use authenticated --setup clone or fresh, with independent writable state and SQLite copy-up completed outside product timing. Record actual strategy and cache policy, not an unsupported cold-cache/reflink claim.
  • Reuse retained historical baselines under the comparison policy amendment above; no old-harness rerun or blanket baseline recollection is required. When a new baseline is necessary for a specific unresolved decision, run only the selected cases with the current harness and align platform, CPU, setup/cache/input and timing boundaries with the candidate.
  • Retain perf.jsonl (header, attempted samples, summary) or verification.json, plus an optional bounded failure.log (default <=1 MiB). No automatic Markdown/HTML/CSV reports, per-sample evidence trees, retained Stores/fixtures/containers, or raw sampler streams. Summaries required for issue closure are derived from compact logs on demand. Prepared cache remains finite and benchmark-owned; reuse [v0.1.3] Optimize benchmark infrastructure: Docker/FUSE, reusable setup, three modes, compact logs #45's cleanup and retention rules.
  • Verification has a hard 59-second end-to-end deadline starting before setup/authentication and ending after cleanup/receipt publication. Reuse existing independent oracles and identity-matched proofs, explicitly report omissions, and return PASS/FAIL/TIMEOUT/INCOMPLETE. Unsupported long proofs are not shortened or relabeled passing. No unchanged automatic retries or withdrawn exhaustive Phase 1 replay.
  • Investigation and independent source edits may run concurrently after handoff. Serialize resource-sensitive builds, preparation, verification and performance across both tracks, not merely within each output directory. Use one coordinator for measurement scheduling and shared-file integration; obey aggregate <=8 participating CPUs and stricter applicable family limits.

Execution discipline

Use #38's two-track ownership and single measurement schedule. Reuse the relevant #45 smoke/setup/proof results and finite prepared cache; record slow or blocked cases explicitly. Each optimization hypothesis starts with one selected case/seed or diagnostic size pair, one predicted counter, and one small shared change. Scale to the original full curve only after the route is correct and the change is useful. Preserve exact source identities for failed/superseded attempts in compact logs. No repeated passing suites, concurrent timed workloads, extra permanent artifact trees or withdrawn exhaustive verification.

This issue becomes successor optimization work after #45's completed handoff. Updating it does not start execution, create a task, close siblings, authorize release publication or launch #39 recovery.

Activity

  1. yifanxuaaa commented on Sep 4, 2026

    @yifanxuaaa
    ContributorAuthor

    Implementation checkpoint, not terminal PASS. Local codex/issue38-four-families starts at46de7d429. Repaired the spilled-candidate ordering MissingObject defect and sequential capture reuse; transferred ordinary Workspace admission to existing owned-page consumption with exact selected physical delivery and checked publication.

    Candidate2 high-pair product seconds (seeds1,2,3):100MiB=[0.462784042,0.452368126,0.470105250];500MiB=[2.446243751,2.359509918,2.554452293]. Normalized median growth1.057185870 FAILS the<=1gate; all samples are retained and within15s. The500MiB exact-current canonical/FUSE verifier passed in8.940959s with zero skipped current bytes/bodies and cleanupPASS; its explicit full_canonical_census_performed=false remains visible. This is not an exhaustive object-storage census.

    The next small root fix retains the existing inline contiguous spool representation for sequential appends instead of creating100/500 piece-tree nodes; regression tests preserve fragmented/overlapping/sparse/limit behavior. A bounded existing offset index now avoids redundant spool framing reads. Candidate3 is building; no final12-sample qualification or acceptance completion is claimed. Compact evidence: target/issue38-evidence/candidate2/; progress: docs/roadmap/0.1/0.1.3/issue38-progress.md. Nothing pushed, merged or released.

    Subsequent user-authorized amendment received from the side conversation: the current median gate is strictly<1.10, described as bounded scaling overhead. The<=1.00 assessments above remain historical strict-gate outcomes, not the amended status. The candidate2 payload/CASunique high-pair medians clear the amended threshold; CASunique seed3~1.134 remains visible. All114 final samples, other curves, verification, resource/runtime and cleanup requirements remain unchanged; no automatic child closure.

  2. yifanxuaaa commented on Sep 5, 2026

    @yifanxuaaa
    ContributorAuthor

    Frozen Docker-owned qualification completed under the user's final strictly<1.25 normalized-median gate. Earlier <=1.00/<1.10 misses remain in the report and raw receipts. This is a25% scaling-overhead allowance, not proportional-scaling or statistical-confidence evidence.

    Local commit 0dc2565f0d3f2ff6271fc743bf4a904b7267b58f (not pushed/merged). Measured source seal eb9b79cd87e6d655e26cad37b69327c70379e831b4d61fb4dd49838795d6fd25, product 20aee20d96b09e0cd1d934b751a1e0370523987adfabc220aa63f283bb7a7d9e, image sha256:f66ce7083572d8c81d43948f5308cc4b96080f548e5a796b57af8d3b40c2c988. Full distributions, historical-approximate comparisons, phase/resources, all30ratios and receipt links: /Users/yifanxu/Ephemeral-AI-Lab/layerfs/docs/roadmap/0.1/0.1.3/issue38-candidate3-docker-results.md; compact evidence under target/issue38-evidence/candidate3/.

    All samples retain actual work and independent setup, remain below15product seconds with zero swap/OOM, and report cleanupPASS. Runtime guards enforce Mac orchestration+LinuxDocker, no Docker data mounts/volumes/socket, and real FUSE/Exec where applicable. Workspace samples use independently writable verified copies of closed masters; native imports reuse source fixtures and initialize fresh outputs. No sample containers remain; preparedcache8entries/1,282,396,850data bytes.

    Required checks:289 native tests passed,0failed,1preexistingignored in Docker;2-test-job/2CPU warmphase96s passes the unchanged120s gate. Formatting and Clippy(warnings denied)pass. The prior1-job177s scheduling failure remains retained. Four final representative verifiers total44.786109666s, maximum16.128966208s. Their no-reuse current-state mode reads all selected current file bytes through canonical reads and FUSE, verifies namespace/metadata/aliases and applicable independent CAS/CDC transcripts; exhaustive typed-object/storage census remains explicitly omitted. No#39 recovery, #40 namespace performance success or release is claimed.

    Case Seed1s Seed2s Seed3s Median s Mean s Min–max s
    payload-create-1m-compact-v2 0.015796376 0.012608416 0.016124625 0.015796376 0.014843139 0.012608416–0.016124625
    payload-create-10m-compact-v2 0.061246375 0.053728376 0.054880708 0.054880708 0.056618486 0.053728376–0.061246375
    payload-create-100m 0.428077750 0.569551709 0.428373834 0.428373834 0.475334431 0.428077750–0.569551709
    payload-create-500m 2.368766125 2.445822585 2.286476000 2.368766125 2.367021570 2.286476000–2.445822585
    Curve Adjacent tiers Normalized median Corresponding seed ratios Current gate
    payload-create 1→10 0.347425941 0.387724/0.426131/0.340353 PASS <1.25
    payload-create 10→100 0.780554496 0.698944/1.060058/0.780554 PASS <1.25
    payload-create 100→500 1.105934087 1.106699/0.858859/1.067514 PASS <1.25

    The complete frozen child scope passes the user-amended gate. Host-owned Store migration follows separately in the existing parent task; these Docker-owned receipts are not relabeled host evidence. Smaller historical points use the explicit comparison classification in the report; no artificial slowdowns were added and causal attribution across topology differences remains limited.

  3. yifanxuaaa commented on Sep 5, 2026

    @yifanxuaaa
    ContributorAuthor

    Additional host-owned Store qualification PASS

    This child remains completed for the separately preserved frozen Docker campaign. Its host follow-up now also passes all required family samples/seeds/scaling/work/runtime/resource checks. Selected separate500-tier proof payload-create-500m, seed3:PASS in7.224484500s, including cleanup; current-byte/metadata/CAS/CDC coverage and the exhaustive-census omission remain explicit.

    Host-owned SQLite follow-up is complete in the same task/worktree, with local commit 95508a3d72c5e3f6ae7796938c773f52d5def45e after frozen product commit 0dc2565f0d3f2ff6271fc743bf4a904b7267b58f. No push, merge or release was performed.

    The host owns SDK/coordinator, Workspace capture/spool and embedded SQLite. Linux containers retain the authenticated daemon, workload and real FUSE. Existing ContainerBinding/ProxyHost is reused, with no new transport/HTTP/data mounts. Product seal remains 20aee20d96b09e0cd1d934b751a1e0370523987adfabc220aa63f283bb7a7d9e. Qualified host source f1eb7c6f3dd1ebd31f639e888953be523c842ad21c8d77d45b1d7c37cef42181; Linux image source remains eb9b79cd87e6d655e26cad37b69327c70379e831b4d61fb4dd49838795d6fd25. Final delivered host source 69cf6ec212bed950f89b25ec547e6713f33e4dadc1fc19c68e9f70c534b5fa5f has only the verifier lint simplification and final harness identity/output guards after qualification; performance operations/product code are unchanged, and a final small performance/verifier check passed. Old receipts retain their actual identities.

    Host resources are separate: container2 CPUs/2 GiB/no swap; host CPU is not container-capped and native initialization uses up to8 existing workers. Host system SQLite3.51.0 and host page-cache ownership differ from the Docker reference. These are matched product/case/seed/timer/work comparisons with a changed topology/resource scope, not pure-efficiency claims. Workspace reuse500 is5.12% slower on host; the <1.25 scaling gate is not a topology-regression allowance.

    Protected closed SQLite masters and native sources live under ignored project-local benchmark-results/host-store/. Independent writable byte copies, fsync, quick_check, matching starting hashes/distinct inodes, unchanged master checks, actual miss→hit across executor changes and bounded ownership-checked eviction passed. Live SQLite sidecars are rejected. Native imports always create fresh output Stores. Final inventory: zero owned containers, empty host samples,8 host cache entries/509,078,386 bytes;8 Docker prepared entries retained. Cache is evictable, not a durable backup; ordinary caller-selected product Store paths are outside cleanup.

    Final checks:20 shared tests, formatting, strict benchmark Clippy, host release build and host v5 migration/staging/publication failure-injection check PASS. Existing live SDK host proof passed post-attach failure/disconnect cleanup; payload verification proves same-session reads/writes after two full-status Commits, returned/published heads, bytes/mode, fresh projection and clean End. Frozen product Linux checks remain289 PASS/one preexisting ignored,96s warm suite, strict Clippy PASS.

    Full per-seed distributions, all normalized ratios, phase/Exec/Commit/transport/copy/admission/host-vs-container resource tables and evidence links are committed locally in docs/roadmap/0.1/0.1.3/issue38-host-store-results.md; frozen Docker evidence remains separately recorded in issue38-candidate3-docker-results.md. Compact host receipts are in benchmark-results/host-store/results/. The initial NotReady verifier failure is preserved; it was a harness owner-lifetime problem corrected by retaining one daemon binding across Store close/reopen. The independent verifiers check all selected current bytes/namespace/metadata/aliases and applicable CAS/CDC transcripts, explicitly not an exhaustive typed-object/storage census.

    Family topology comparison

    Case Actual work bytes Host seeds1 / 2 / 3 s Docker median s Host median s Host/Docker
    payload-create-1m-compact-v2 1048576 0.025027875 / 0.024693665 / 0.025669625 0.015796376 0.025027875 1.584406
    payload-create-10m-compact-v2 10485760 0.062607749 / 0.057788541 / 0.062064000 0.054880708 0.062064000 1.130889
    payload-create-100m 104857600 0.529123250 / 0.408488916 / 0.396174500 0.428373834 0.408488916 0.953580
    payload-create-500m 524288000 2.402669499 / 1.881036958 / 1.899561791 2.368766125 1.899561791 0.801920
  4. added 2 commits that reference this issue on Nov 15, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    v0.1.3LayerFS v0.1.3 checkpoint and delivery

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions