HTTP: 503 Service Unavailable
Last verified
503 Service UnavailableWhat it means
503 means the server is temporarily unable to handle the request, usually because of overload or maintenance. RFC 9110 lets it send Retry-After to say when to come back, exactly as with 429.
Automation sees 503 in three situations. First, genuine maintenance or overload. Second, some sites and CDNs use 503 as a soft rate limit or for a bot challenge page: if the body is an HTML challenge rather than a maintenance notice, the site wants an interactive browser. Third, Google's crawling documentation describes 503 as the recommended way for a site to ask crawlers to slow down temporarily, so polite crawlers treat it as a back-off signal.
What is not wrong: the URL. The same request should succeed later if you wait.
What it looks like
HTTP/1.1 503 Service Unavailable
Retry-After: 120Why it happens
- Scheduled maintenance or a deploy.
- Server overload from traffic spikes, including your own concurrency.
- A rate-limiting layer that responds with 503 instead of 429.
- A bot-protection challenge page delivered with a 503 status.
How to fix it
- Honor Retry-After if present; otherwise back off exponentially with jitter.
- Lower request rate and concurrency for that host.
- Inspect the body: a challenge page means the site restricts automated clients; check its terms and look for an API.
- Check the site's status page.
- Resume gradually after recovery instead of replaying the whole backlog at once.
How to stop it recurring
Use a shared rate limiter per host, honor Retry-After on both 429 and 503, and stop the job if 503s persist instead of retrying forever. Schedule large crawls outside the site's peak hours, and treat a 503 challenge page as a policy signal: look for an API or data feed rather than trying to get past it.
Official documentation: developer.mozilla.org/en-US/docs/Web/HTTP/Reference/Status/503. See all HTTP error messages.
Related errors
Definitions
- CAPTCHACAPTCHA (Completely Automated Public Turing test to tell Computers and Humans Apart) is a challenge-response m
- Cloudflare Bot ManagementCloudflare Bot Management is the set of Cloudflare features (Bot Fight Mode, Super Bot Fight Mode and the Ente
- DataDomeDataDome is a commercial bot-protection service that scores every request to a website, app or API in real tim
- HUMAN Security (PerimeterX)PerimeterX, now HUMAN Security's Bot Defender after a 2022 merger, is a bot-protection service that scores eve
- User AgentA user agent is a string that browsers send with every HTTP request identifying the browser type, version, and
- AI Web ScraperAn AI web scraper is an AI-powered data extraction tool that understands page structure, identifies relevant d
- Data ScrapingData scraping is the broad practice of programmatically extracting data from any digital source — websites, ap
- Web CrawlerA web crawler (also called a spider or bot) is a program that systematically browses the internet by following
Quick answers
What does "503 Service Unavailable" mean in HTTP?
503 means the server is temporarily unable to handle the request, usually because of overload or maintenance. RFC 9110 lets it send Retry-After to say when to come back, exactly as with 429. Automation sees 503 in three situations. First, genuine maintenance or overload.
How do I fix "503 Service Unavailable"?
1. Honor Retry-After if present; otherwise back off exponentially with jitter. 2. Lower request rate and concurrency for that host. 3. Inspect the body: a challenge page means the site restricts automated clients; check its terms and look for an API. 4. Check the site's status page. 5. Resume gradually after recovery instead of replaying the whole backlog at once.
Retries and timeouts without writing the loop
Autonoly's HTTP Request node and browser steps have per-node retries, timeouts and continue-on-fail, so a transient 429 or 503 can back off instead of failing the run. It has no proxies or IP rotation: if a site blocks automated traffic, that block applies to Autonoly too.
Also see Error handling in workflows.