merge: dev → main (fee-compare improvements)

.github/workflows/prod-deploy.yml

@@ -107,10 +107,16 @@ jobs:
       # the next step auto-rolls back the prod alias to the previous good
       # deploy. This catches the class of bug that bled SEO on 2026-06-25
       # (bench page 404 + noindex while still in the sitemap).
+      #
+      # Sitemap fetched from the deployment URL (bypasses s-maxage=3600
+      # CDN cache — was serving the previous deploy's sitemap for up to
+      # 1 h, causing false rollbacks). Pages checked against the prod
+      # domain (second arg) so the noindex injected on Vercel preview
+      # URLs does not false-fail the per-page checks.
       - name: Sitemap indexability smoke
         id: smoke
         continue-on-error: true
-        run: node scripts/sitemap-smoke.mjs https://openchainbench.com
+        run: node scripts/sitemap-smoke.mjs ${{ steps.deploy.outputs.url }} https://openchainbench.com
 
       - name: Auto-rollback on smoke failure
         if: steps.smoke.outcome == 'failure'

benchmarks/acala-rpc.yml

@@ -0,0 +1,136 @@
+# OpenChainBench. Bench No 248
+
+slug: acala-rpc
+number: "248"
+title: Fastest free Acala RPC, live no-key endpoint latency
+seo_title: "Fastest free Acala RPC 2026"
+seo_description: "{{best_name}} leads free Acala RPC at {{best_p50}} (block height p50, 24h). 3 providers measured every 60s from 3 regions."
+subtitle: HTTP round-trip latency for Substrate chain_getHeader queries against every available public Acala endpoint, audited every 60 seconds from 3 regions.
+
+category: RPCs
+status: live
+metric: RPC latency
+unit: ms
+higher_is_better: false
+
+seo_intro: |
+  Acala is the DeFi hub of the Polkadot ecosystem, a Polkadot parachain providing decentralized stablecoin (aUSD), liquid staking (LDOT), and an EVM-compatible execution environment. Acala uses the Substrate framework and exposes a standard Substrate JSON-RPC interface; block height is queried via chain_getHeader. Public endpoints are available without an API key from Dwellir and the Acala official RPC cluster (aca-api.network). Every provider was live-verified with consecutive block-height probes at launch.
+
+abstract: |
+  Per-chain member of the RPC latency cluster, extended to Acala.
+  We measure the round-trip latency of a Substrate chain_getHeader query against
+  every available public Acala endpoint that sustains continuous probing:
+  3 providers at launch, every 60 seconds, from us-east, eu-west and
+  Singapore. Each provider is probed via a JSON-RPC POST (chain_getHeader)
+  from which the block number is extracted.
+  The harness classifies every response (ok / http_err / jsonrpc_err / stale / timeout)
+  with a Polkadot-scaled staleness gap (40 blocks, around 8 min at 12 s/block).
+  The cross-chain view lives on the parent rpc-capabilities benchmark;
+  this page is the Acala-scoped answer with per-region breakdowns as
+  a first-class dimension.
+
+methodology:
+  - "Cadence: every 60 seconds per provider, from each of 3 probe regions (us-east Virginia, eu-west Amsterdam, sgp Singapore). Headline p50/p90/p99 aggregate across all 3 regions via Prometheus avg(quantile_over_time(...)); per-region breakdowns are first-class on this page via the region tabs."
+  - "Payload: {\"jsonrpc\":\"2.0\",\"id\":<rotating>,\"method\":\"chain_getHeader\",\"params\":[]}. The result contains a number field (hex-encoded block number)."
+  - "Latency: client-side round-trip delta in milliseconds, exposed as both a gauge and a histogram (buckets 50 ms to 10 s), so percentiles are computed via Prometheus quantile_over_time over the last 24 hours."
+  - "Call-result classification: ok (parsable block height returned), http_err, jsonrpc_err, stale (block more than 40 behind the cross-provider tip), timeout."
+  - "This page is part of the per-chain RPC cluster derived from the cross-chain rpc-capabilities benchmark; the identical harness, cadence and exclusion rules apply on every chain."
+  - "Chain scope: every query on this page is pinned to chain=acala. Provider coverage at launch: 3 endpoints (Dwellir, Acala RPC-0, Acala RPC-2)."
+
+findings:
+  - "{{best_name}} currently leads Acala RPC at {{best_p50}} (chain_getHeader p50, 24h) across 3 measured providers."
+
+faq:
+  - q: "What is the fastest free Acala RPC right now?"
+    a: "{{best_name}} currently leads at {{best_p50}} (Acala block height p50 over the last 24h), measured against 3 providers probed every 60 seconds from us-east, eu-west and Singapore. The leaderboard re-sorts continuously against fresh Prometheus samples."
+  - q: "Which Acala RPC endpoints work without an API key?"
+    a: "3 endpoints sustain continuous keyless probing at launch: Dwellir (acala-rpc.n.dwellir.com), Acala RPC-0 (acala-rpc-0.aca-api.network) and Acala RPC-2 (acala-rpc-2.aca-api.network). Every listed endpoint was live-verified before inclusion."
+  - q: "Does the fastest Acala RPC change by region?"
+    a: "Often. Dwellir is Europe-focused while the Acala official nodes have fixed infrastructure. Rankings shift depending on your geographic origin. The region tabs re-scope every number to a single origin."
+  - q: "How does Acala RPC latency compare to other Polkadot parachains?"
+    a: "All Polkadot parachains use the same Substrate JSON-RPC probe (chain_getHeader), making latency directly comparable across the per-chain RPC cluster. The parent rpc-capabilities benchmark shows the cross-chain view."
+
+source: https://github.com/ChainBench/OpenChainBench/tree/main/harnesses/rpc-capabilities
+
+prometheus:
+  window: 24h
+  freshness_metric: rpc_latency_milliseconds
+
+rank_matrix_query: avg by (provider, region) (ocb:rpc_latency_milliseconds:p50_24h{chain="acala"})
+
+dimensions:
+  region:
+    - { value: all,     label: All regions }
+    - { value: us-east, label: US-East }
+    - { value: eu-west, label: EU-West }
+    - { value: sgp,     label: Singapore }
+
+providers:
+  - slug: dwellir
+    name: Dwellir
+    tag: Dwellir public Acala RPC, no API key required
+    formula: "50th percentile over 24h of client-side round-trip latency (ms) for a chain_getHeader POST sent every 60s from 3 regions to acala-rpc.n.dwellir.com."
+    queries:
+      p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="dwellir", chain="acala"})
+      p90: avg(ocb:rpc_latency_milliseconds:p90_24h{provider="dwellir", chain="acala"})
+      p99: avg(ocb:rpc_latency_milliseconds:p99_24h{provider="dwellir", chain="acala"})
+      mean: avg(ocb:rpc_latency_milliseconds:mean_24h{provider="dwellir", chain="acala"})
+      success: sum(ocb:rpc_call:ok_rate_24h{provider="dwellir", chain="acala"}) / sum(ocb:rpc_call:rate_24h{provider="dwellir", chain="acala"})
+      sample_size: sum(ocb:rpc_call:increase_24h{provider="dwellir", chain="acala"})
+      series: avg(avg_over_time(rpc_latency_milliseconds{provider="dwellir", chain="acala"}[1h]))
+      regions:
+        - region: us-east
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="dwellir", chain="acala", region="us-east"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="dwellir", chain="acala", region="us-east"}[1h])
+        - region: eu-west
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="dwellir", chain="acala", region="eu-west"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="dwellir", chain="acala", region="eu-west"}[1h])
+        - region: ap-southeast
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="dwellir", chain="acala", region="sgp"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="dwellir", chain="acala", region="sgp"}[1h])
+
+  - slug: acala-rpc0
+    name: Acala (RPC-0)
+    tag: Acala official public RPC node 0 (aca-api.network), no API key required
+    formula: "50th percentile over 24h of client-side round-trip latency (ms) for a chain_getHeader POST sent every 60s from 3 regions to acala-rpc-0.aca-api.network."
+    queries:
+      p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc0", chain="acala"})
+      p90: avg(ocb:rpc_latency_milliseconds:p90_24h{provider="acala-rpc0", chain="acala"})
+      p99: avg(ocb:rpc_latency_milliseconds:p99_24h{provider="acala-rpc0", chain="acala"})
+      mean: avg(ocb:rpc_latency_milliseconds:mean_24h{provider="acala-rpc0", chain="acala"})
+      success: sum(ocb:rpc_call:ok_rate_24h{provider="acala-rpc0", chain="acala"}) / sum(ocb:rpc_call:rate_24h{provider="acala-rpc0", chain="acala"})
+      sample_size: sum(ocb:rpc_call:increase_24h{provider="acala-rpc0", chain="acala"})
+      series: avg(avg_over_time(rpc_latency_milliseconds{provider="acala-rpc0", chain="acala"}[1h]))
+      regions:
+        - region: us-east
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc0", chain="acala", region="us-east"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc0", chain="acala", region="us-east"}[1h])
+        - region: eu-west
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc0", chain="acala", region="eu-west"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc0", chain="acala", region="eu-west"}[1h])
+        - region: ap-southeast
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc0", chain="acala", region="sgp"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc0", chain="acala", region="sgp"}[1h])
+
+  - slug: acala-rpc2
+    name: Acala (RPC-2)
+    tag: Acala official public RPC node 2 (aca-api.network), no API key required
+    formula: "50th percentile over 24h of client-side round-trip latency (ms) for a chain_getHeader POST sent every 60s from 3 regions to acala-rpc-2.aca-api.network."
+    queries:
+      p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc2", chain="acala"})
+      p90: avg(ocb:rpc_latency_milliseconds:p90_24h{provider="acala-rpc2", chain="acala"})
+      p99: avg(ocb:rpc_latency_milliseconds:p99_24h{provider="acala-rpc2", chain="acala"})
+      mean: avg(ocb:rpc_latency_milliseconds:mean_24h{provider="acala-rpc2", chain="acala"})
+      success: sum(ocb:rpc_call:ok_rate_24h{provider="acala-rpc2", chain="acala"}) / sum(ocb:rpc_call:rate_24h{provider="acala-rpc2", chain="acala"})
+      sample_size: sum(ocb:rpc_call:increase_24h{provider="acala-rpc2", chain="acala"})
+      series: avg(avg_over_time(rpc_latency_milliseconds{provider="acala-rpc2", chain="acala"}[1h]))
+      regions:
+        - region: us-east
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc2", chain="acala", region="us-east"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc2", chain="acala", region="us-east"}[1h])
+        - region: eu-west
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc2", chain="acala", region="eu-west"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc2", chain="acala", region="eu-west"}[1h])
+        - region: ap-southeast
+          p50: avg(ocb:rpc_latency_milliseconds:p50_24h{provider="acala-rpc2", chain="acala", region="sgp"})
+          series: avg_over_time(rpc_latency_milliseconds{provider="acala-rpc2", chain="acala", region="sgp"}[1h])