diff --git a/MANAGED_TARGET.md b/MANAGED_TARGET.md index 08e52a5..c8e1f85 100644 --- a/MANAGED_TARGET.md +++ b/MANAGED_TARGET.md @@ -1,5 +1,49 @@ # Fresh managed-test target contract (isolated acceptance, not production approval) +## Declarative HTTPS resource checks v1 (check-only) + +Inside the provisioned owned managed target, with the tracked example available: + +```sh +zapret-autotune check --config resource-check.example.json +``` + +This is not a standalone host checker for arbitrary external proxies: owned-target, +SS and configured wire-DNS readiness/ownership guards must pass before probes. +Version1 supports `kind: http`, GET/HEAD on HTTPS **DNS-hostname** URLs, +TLS1.2/1.3/both, timeout and expected HTTP statuses, MIME content type and +minimum/maximum bytes. Outcomes identify stable configured service/resource IDs +and report HTTP status, bytes and latency. URL/header secrets come from env via +private stdin, not argv; keep the config and env private when sensitive. + +All resolved IPs are policy-checked before numeric CONNECT pinning; TLS verifies +the original hostname/SNI. Redirects and system-DNS fallback are forbidden. +Private destinations require explicit opt-in **and** CIDR restriction. IPv6 +literal URLs and address failover are not supported in v1. Image/media decoding, +Range and suffix-scope qualification are not generic HTTP-check features. + +Schema/loader and executable checks have unit and isolated remote acceptance for +source `4d7c4b33c36186d43ea89fd24354b1834b21692c`, image +`sha256:be5a1c45c6d40b390a29bfc9a5ab3975c3a8e0c015b98a524696312807dad9a4`. +The orchestrator supplied verified private evidence, not reread by this docs-only +update; see [PROGRESS.md](PROGRESS.md) for gates and cleanup. The positive private +TLS fixture trusted a **staging-only test CA**, never bypassed TLS or changed +production CAs. The existing DNS backend and earlier DoH outage evidence are reused, +not a new DoH implementation or new outage test here. + +**Acceptance scope:** staging YouTube monitoring was SIGSTOP-isolated from checks, +then SIGCONT-resumed before graceful stop (exit0, owned cleanup, foreign baseline +unchanged). A prior managed monitor exit78 remains a runtime caveat, not proof of +whole-controller resilience. LIVE11082 was already stopped, not upgraded/restarted; +old SS1080/Xray were untouched. Historical LIVE observations below are not current +production availability. No network operations or deployment accompany this update. + +Check-only neither selects/applies strategies nor writes strategy DB entries or +declares full-service READY. Generic periodic controller/config/strategy DB, +plugins/autotuning, maximum strategy collection, safe PID1 BLOCKED stay-alive and +Prometheus metrics remain TODO. Three-CDN discovery/qualification is a YouTube +adapter contract, not a universal arbitrary-service requirement. + ## Managed DoH live acceptance (2026-10-13): DNS verified, NOT READY Supplied verified reports bind source `544dda7c85965fc3200de3af17fca9ef9999b2ce` @@ -95,8 +139,9 @@ whole-service/avatar acceptance remains open. The 2026-10-12 measured results below are historical, not the new factory evidence. New images still require their own source/image binding and guarded apply/rollback acceptance. Maximum strategy collection/search diversity and PID1 global-failure stay-alive remain unfinished future -increments. Arbitrary declarative service configuration is future work; no new -generic policy framework is supplied. See [GOOGLEVIDEO_DISCOVERY.md](GOOGLEVIDEO_DISCOVERY.md). +increments. Declarative check-only configuration is accepted above; generic +controller integration remains future work. No new generic policy framework is +supplied. See [GOOGLEVIDEO_DISCOVERY.md](GOOGLEVIDEO_DISCOVERY.md). `managed init|reconcile|status|active-ids|apply` is exclusively for a **new dedicated container** which owns its entire config. The existing/dev/production container must refuse diff --git a/PROGRESS.md b/PROGRESS.md index f322e17..171c62a 100644 --- a/PROGRESS.md +++ b/PROGRESS.md @@ -4,6 +4,14 @@ ## Текущий статус +### Declarative resource-check acceptance (2026-10-13): check-only PASS, not controller acceptance + +- Supplied orchestrator-verified remote evidence binds source **`4d7c4b33c36186d43ea89fd24354b1834b21692c`** to image **`sha256:be5a1c45c6d40b390a29bfc9a5ab3975c3a8e0c015b98a524696312807dad9a4`**. Private report: `/home/agent/zapret2-autotune-resources-4d7c4b3/field-private/final.safe.json`; privileged-read by the orchestrator, **not independently reread here**. Reported `checker_gates_passed=true`; later documentation commits are not the tested source. No new network checks or deployment accompany this update. +- JSON v1 schema/loader and executable HTTPS GET/HEAD checks are unit-verified and now remotely accepted: public GET/HEAD on TLS1.2/1.3; private opt-in plus CIDR with a staging-only trusted test CA, original TLS/SNI hostname and numeric CONNECT pin. No `--insecure`, production trust-store pollution or DoH implementation change. Checks reuse the owned SS/configured wire-DNS backend; the earlier DoH outage acceptance recorded below remains separate evidence. +- Default/private-policy, mixed DNS answers, metadata/link-local/ULA/loopback denials made **zero destination connections**; rebinding used the first pinned address and denied the later answer. Ownership-negative cases made zero DNS/resource probes. Cancellation exited1 with joined workers in **0.38s**; the secret sentinel was absent from argv, output and persisted artifacts. +- **Check-only isolation:** staging YouTube monitor was SIGSTOP-isolated during checks, then SIGCONT-resumed **before** graceful stop; reported `monitor_resumed=true`, Docker exit0. Cleanup left zero owned host PIDs, no owned namespace or temporary containers/networks, and an unchanged foreign baseline. An earlier existing managed monitor **exit78** contributed before isolation: this does **not** prove whole-controller resilience, generic recovery or production availability. LIVE11082 was already **stopped** before the task and was **not upgraded or restarted**; old SS1080/Xray were untouched. Earlier LIVE snapshots below are historical, not current uptime claims. +- Remaining P0: generic periodic controller integration, config/strategy DB integration, plugins/autotuning. Maximum strategy collection and PID1 safe stay-alive/BLOCKED mode remain TODO; Prometheus is P1 TODO, not an implemented endpoint. Three-CDN qualification belongs to the YouTube adapter, not every arbitrary service. See [MANAGED_TARGET.md](MANAGED_TARGET.md#declarative-https-resource-checks-v1-check-only) and the tracked [resource-check.example.json](resource-check.example.json). + ### Managed DoH acceptance (2026-10-13): live DNS path and staged outage verified, NOT READY - Managed target defaults to the stdlib Go UDP/TCP loopback DoH forwarder; PID 1 validates owned readiness before ss-server `-d 127.0.0.1:5353`. SOCKS remote-hostname traffic and managed `socks5h` checks share this path. Provider URL/bootstrap IPs/timeout are private-env configurable; explicit native mode alone preserves legacy DNS, never automatic fallback. @@ -31,7 +39,7 @@ All **18 post-warmup transfers** returned HTTP206, valid Range `0..65535`, 65536 bytes, MP4 `ftyp`, and valid SSL verification. Warm failures are separate, not cold qualification. Fresh version-1 `owned-nfq-media-range-v1` attestation in Adaptive mode retained actual host/endpoints, source IDs and native-owner Lua binding; proof checked at `2026-10-13T08:08:28.752Z`, with three roots and three prefixes. Native request statistics were not persisted: manual request count is unknown, not zero or an invented total. - The operator automatically **PROMOTED** one newly qualified Adaptive candidate10: generation9→10, active IDs `[1,3,9]`→`[1,9,10]`, scope `experimental_googlevideo_suffix`→`suffix`, current config hash prefix `dc997ac91…`. Common hostlist `googlevideo.com`, frontend1 and thumbnail9 were retained. Qualification and fresh deployed checks used the transaction-bound apply safety path (asset/frontend hooks have source test coverage). This qualifies the measured representative three-CDN composite profile, not every Googlevideo host, cold behavior, or a maximum strategy catalogue. The shared profile remains two native strategies with per-host learning. - Final reported snapshot at **08:21:48**: container running and monitor alive; whole-service health still false / **DEGRADED, NOT READY**. Probe failures reached405 at 08:20:37; their avatar-specific cause was not independently measured in this phase. The prior day's avatar timeouts are historical, not today's diagnosis. The orchestrator's fresh owned-SOCKS check returned `0500`; role-owned `socks5h` uses native SS DNS `8.8.8.8`. The passive SSH watch session vanished after about ten minutes: no normally completed 18-minute watch or interruption of the controller is claimed. -- Remaining: actual avatar/service health qualification, maximum strategy collection and broader search diversity, and controller security/global-failure stay-alive policy (current safety guards remain fatal). Per-service failure isolation and arbitrary declarative service configuration are future requirements, not a new plugin framework. No further build, probes or cooldown changes accompany these docs. +- Remaining: actual avatar/service health qualification, maximum strategy collection and broader search diversity, and controller security/global-failure stay-alive policy (current safety guards remain fatal). Generic periodic per-service failure isolation/controller integration remain future requirements; declarative check-only configuration is accepted above, not a new plugin framework. No further build, probes or cooldown changes accompany these docs. ### Autonomous three-CDN source behavior (2026-10-13) @@ -59,8 +67,8 @@ unimplemented describe historical RCs, not this source; historical measurements are not the new owned factory evidence. - Maximum strategy collection and PID1 global-failure stay-alive are still future - increments; broad maximum-profile live staging is deferred. Arbitrary declarative - service configuration remains a future note, not a new framework. Details: + increments; broad maximum-profile live staging is deferred. Declarative checks + are accepted above; generic service controller integration remains future work. Details: [GOOGLEVIDEO_DISCOVERY.md](GOOGLEVIDEO_DISCOVERY.md). ### Final managed RC field acceptance (2026-10-12): DEGRADED, not READY