Steem API Health — Daily Report 2026-09-09

in #steem12 days ago

Steem API Health · 2026-09-09

On 2026-09-09 (UTC) the monitor tracked 12 Steem API nodes with a global uptime of 91.41% across 16,944 measurements. Median latency across all nodes: 184 ms. Today's full picture and notable patterns below.

All measurements are taken from a single European location (Germany). Latency to nodes hosted outside Europe will naturally be higher than for users connecting from those regions. Want to contribute measurements from your region? See the participation block below.

Observations

  • Longest outage today: steemd.blazeapps.org, 1412 consecutive failed minutes.
  • Fleet trend: median latency regressed by 33 ms vs. yesterday (+33 ms; now 184 ms).
  • Fastest node: api.moecki.online at 76 ms average. That is 249 ms below the fleet average of 325 ms across 11 nodes.
  • Slowest node: steem.justyy.com at 913 ms average. That is 588 ms above the fleet average of 325 ms.
  • Consistent leader: api.campingclub.me has been in the top 3 by latency every day this week.
  • Consistent leader: api.justyy.com has been in the top 3 by latency every day this week.
  • Consistent leader: api.moecki.online has been in the top 3 by latency every day this week.
  • Consistent laggard: api.steemit.com has been in the bottom 3 by latency every day this week.
  • Consistent laggard: api.steemyy.com has been in the bottom 3 by latency every day this week.
  • Consistent laggard: steem.justyy.com has been in the bottom 3 by latency every day this week.

Nodes

NodeRegionUptimeAvgp50p95p99ErrorsError classes
api.moecki.onlineeu-central100.00 %76 ms79 ms110 ms126 ms0
steemd.steemworld.orgeu-central100.00 %101 ms101 ms106 ms122 ms0
api.campingclub.measia100.00 %102 ms101 ms143 ms162 ms0
api3.justyy.comasia100.00 %152 ms153 ms196 ms241 ms0
steem.senior.workers.devglobal100.00 %184 ms179 ms269 ms321 ms0
api.steem.fansasia99.93 %327 ms234 ms618 ms652 ms1rpc_error ×1
api2.steemyy.comglobal99.93 %456 ms302 ms945 ms1035 ms1rpc_error ×1
api.justyy.comus-west99.86 %78 ms70 ms96 ms317 ms2rpc_error ×2
steem.justyy.comus-central99.86 %913 ms838 ms1142 ms3452 ms2http_5xx ×1, rpc_error ×1
api.steemyy.comunknown98.73 %557 ms481 ms812 ms3339 ms18http_4xx ×15, rpc_error ×2, timeout ×1
api.steemit.comus-east98.58 %627 ms553 ms872 ms2524 ms20rpc_error ×1, timeout ×19
steemd.blazeapps.orgeu-central0.00 %1412connect_error ×1412

Latency distribution

Across 15,488 successful measurements today: 56.2 % under 200 ms, 71.2 % under 500 ms, 98.4 % under 1 000 ms. (1.6 % were slower than 1 second.)

Time-of-day pattern

Best hour today: 21:00–22:00 UTC (avg latency 286 ms, 648 measurements). Worst hour: 11:00–12:00 UTC (avg latency 358 ms, 649 measurements).

Error pattern

Most common error today: connect_error (97.0 %), followed by timeout (1.4 %), followed by http_4xx (1.0 %) (of 1,456 total errors).

Biggest outage of the day

steemd.blazeapps.org had a stretch of 1412 consecutive failed minutes today. Full per-tick view: node detail page.

Week over week

Week-over-week: current week 91.26 %, previous week 91.51 % (Δ -0.25 pp).

NodeΔ uptime
steem.justyy.com-1.80 pp
api.steemyy.com-0.80 pp
api.steemit.com-0.61 pp
steem.senior.workers.dev-0.04 pp
api2.steemyy.com-0.02 pp
steemd.blazeapps.org±0.00 pp
steemd.steemworld.org±0.00 pp
api.justyy.com+0.01 pp
api.campingclub.me+0.01 pp
api3.justyy.com+0.04 pp
api.steem.fans+0.08 pp
api.moecki.online+0.17 pp

Best vs worst performance gap

The fastest node was 12.01× faster than the slowest today (76 ms vs 913 ms). Last week the same factor was 20.04× — the gap is narrowing.

30-day reliability ranking

Most reliable

    1. api.campingclub.me — 100.00 %
    1. steemd.steemworld.org — 100.00 %
    1. api3.justyy.com — 99.95 %

Least reliable

    1. steemd.blazeapps.org — 0.00 %
    1. steem.justyy.com — 89.42 %
    1. api.steemyy.com — 98.37 %

Longest unbroken uptime streak: steemd.steemworld.org — 22 days without a single failed tick.

Visual detail

Steem API Health · detail · 2026-09-09

Top: latency distribution per node. Middle: hourly performance. Bottom: cross-region comparison (when multi-source data is available).

Methodology: one condenser_api.get_dynamic_global_properties request per node every 60 seconds, 8-second timeout. Full rules and thresholds in MEASUREMENT-METHODOLOGY.

Want to make these reports more accurate?

Anyone can contribute measurements from their own server. The participant script runs in Docker (3 commands to install) and helps build a global view of node performance.

Contributors get attribution on the Sources page.

Feedback wanted

Have ideas for additional metrics, views, or analyses you'd like to see? Leave a comment below — the report format is still evolving and your input shapes future versions.

Resources


Measurement window: 2026-09-09T00:00:00Z — 2026-09-10T00:00:00Z. Source location: contabo-de-1.