Hex PM Outage History

Hex PM is up right now

Hex PM had 1 outage in the last 2 years totaling 1h 12m of downtime — averaging 0 incidents per month.

There was 1 Hex PM outage since July 29, 2026 totaling 1h 12m of downtime. Each is summarised below — incident details, duration, and resolution information.

Source: https://status.hex.pm

Major July 29, 2026

Outage

Detected by Pingoru
Jul 29, 2026, 06:45 PM UTC
Resolved
Jul 29, 2026, 07:57 PM UTC
Duration
1h 12m
Affected: Hex.pm Website & APIPrivate repository (organizations)Private HexDocs (organizations)
Timeline · 5 updates
  1. investigating Jul 29, 2026, 06:45 PM UTC

    We are currently investigating this issue.

  2. identified Jul 29, 2026, 06:58 PM UTC

    The hex.pm website and API are unavailable. The cause is a surge of automated traffic against package file preview pages, which exhausted memory on the web servers and took them out of rotation. We are applying mitigations now. Public package downloads and hexdocs.pm are not affected. Private package downloads and private documentation that triggers reauthentication will currently fail due to the website and API outage.

  3. monitoring Jul 29, 2026, 07:12 PM UTC

    hex.pm and the API are responding normally again as of 19:11 UTC. The cause was a surge of automated traffic against package file preview pages, which exhausted memory on the web servers. We have blocked that traffic at the edge and raised capacity limits. Package file preview pages are temporarily returning errors while the block is in place.

  4. monitoring Jul 29, 2026, 07:20 PM UTC

    We are continuing to monitor for any further issues.

  5. resolved Jul 29, 2026, 07:57 PM UTC

    hex.pm and the API have been stable since 19:11 UTC, and package file preview pages have been serving normally since 19:36 UTC. The cause was a sustained flood of automated traffic against package file preview pages. Because each of those requests unpacks a package archive, the volume exhausted memory on the web servers and took them out of rotation, which is why the site returned errors. We have put rate limiting in place at the edge and raised capacity limits. The traffic is still arriving and is being rate limited. We are continuing to monitor.

Read the full incident report →