Shadow Outage Ledger: Vendor-Claimed Status
Pricing
from $10.00 / 1,000 run_starts
Run this on a short schedule. Each run reads what the vendor status page claims and measures a public endpoint of the same service from outside. Each check writes one typed event: MATCH, SHADOW_OUTAGE when the endpoint fails and the vendor still claims th
1 to 25 vendors. Each entry is an object with 'name', 'status_json_url', 'probe_url', 'probe_method' and 'expect_status'. 'status_json_url' must be a public Statuspage v2 status document, normally https://status..com/api/v2/status.json, or an Instatus summary document. 'probe_url' must be a public endpoint of the service that needs no credentials; the run measures this URL and compares the result with the vendor claim. 'probe_method' is GET or HEAD, and it defaults to GET. 'expect_status' lists the HTTP status codes that count as a healthy answer, and it defaults to [200]. An entry this Actor cannot read gets its own row and does not stop the run.
![TikTok Profile Posts - Bulk Video Scraper [NO LOGIN] ✅ avatar](https://images.apifyusercontent.com/rYWedKi5oej5A8kulfjYXbmcYFI9BQm3TYwYHdkWFmI/rs:fill:76:76/cb:1/aHR0cHM6Ly9hcGlmeS1pbWFnZS11cGxvYWRzLXByb2QuczMudXMtZWFzdC0xLmFtYXpvbmF3cy5jb20vc0NDU2R3bm1LTUoxUTJzVG4tYWN0b3ItVkpBWGdEbW5qTXdSYWd1U2stNjJSZ1dYeHZJdS1jYW52YS10aWstdG9rLWxvZ28tTUFHek5nb0lIancucG5n.webp){kind=avatar}
[ { "name": "GitHub", "status_json_url": "https://www.githubstatus.com/api/v2/status.json", "probe_url": "https://github.com/robots.txt", "probe_method": "GET", "expect_status": [ 200 ] }, { "name": "Cloudflare", "status_json_url": "https://www.cloudflarestatus.com/api/v2/status.json", "probe_url": "https://www.cloudflare.com/cdn-cgi/trace", "probe_method": "GET", "expect_status": [ 200 ] }, { "name": "npm", "status_json_url": "https://status.npmjs.org/api/v2/status.json", "probe_url": "https://registry.npmjs.org/-/ping", "probe_method": "GET", "expect_status": [ 200 ] }]The interval of the schedule that runs this Actor. The run does not wait; it uses this number to measure the gap since the last check. A gap of more than two intervals means that the schedule missed a run, and each event then says that the ledger has a hole. Set this to the same value as your schedule.
The largest number of probes of one target in one run. A probe that answers as expected ends the repeats, so a healthy vendor costs one request. One short failure is noise, and a repeat keeps the ledger free of a window that opens on it.
Timeout for the status read and for each probe. A probe that reaches this timeout counts as a failure, because a user waits the same time.
The largest number of vendors checked at the same time. A public status page refuses a burst, so a small number is safer.
The largest number of redirects followed for one read or one probe. Every hop is checked again against the private-address rule. A probe stops at the first status code that your 'expect_status' list names, so a buyer who expects 301 measures the redirect itself.
The largest body this Actor reads. A status document larger than the cap is refused. A probe keeps its status code and its latency, because the cap stops the body only.
The named key-value store that carries the ledger between runs. Each run gets its own default store, so the ledger cannot live there. Keep one name for one schedule. Give a second schedule a second name, or the two schedules write over each other. The name must be 3 to 63 characters, and it must use letters, numbers and inner dashes only.
The ledger keeps the open windows and the last vendor claim in the key-value store between runs. A vendor that the input no longer names is kept for this many hours and is then dropped, so the stored state does not grow without a bound. The record never holds more than 200 vendors.