FAIR

Federated and Independent Repositories

Honeylog

Plugin Banner

Honeylog

by secondbreakfastio

Download
Description

Honeylog WordPress Plugin

Multi-file WordPress plugin that batches frontend requests into a JSON queue and flushes to Honeylog.

This client targets self-hosted and conventional managed WordPress. It is not
the complete integration for WordPress VIP, where edge cache hits do not execute
PHP and the uploads filesystem is not a durable local queue. Use
clients/wordpress-vip-s3 for VIP.

Behavior

  • Tracks HTML/XHTML frontend responses regardless of the request Accept header, plus .xml and .txt URLs (excludes wp-admin pages).
  • Keeps frontend responses without an explicit Content-Type, including redirects.
  • Writes each event to a local NDJSON queue file.
  • Serializes queue writes and rotation using a blocking filesystem lock on local storage.
  • Schedules a flush every 60 seconds through WP-Cron; actual execution depends on the cron runner.
  • Sends queued events in batches of up to 100.
  • Warns administrators when automatic WP-Cron execution is disabled.

Plugin Files

  • honeylog-ingestion.php: bootstrap and WordPress hooks
  • includes/honeylog-constants.php: constants + option sanitizers
  • includes/honeylog-batching-utils.php: queue file lock, enqueue/snapshot/requeue/flush
  • includes/honeylog-request-utils.php: request capture + HTTP batch sender
  • includes/honeylog-settings.php: admin settings page
  • includes/honeylog-debug.php: persistent debug log + last status/error helpers

Settings

Go to Settings -> Honeylog:

  • Token
  • Site Domain (optional)

The ingestion endpoint is fixed to https://api.honeylog.io/events.

Debugging

Debug log is written to:

  • wp-content/uploads/honeylog-ingestion/debug.log when available
  • fallback: wp-content/honeylog-ingestion-debug.log

Reliable scheduling and cache coverage

On low-traffic or cached sites, configure a system cron to run due WordPress
cron events every minute. The 60-second schedule is not a delivery deadline.
See docs/clients/wordpress-installation.md for a WP-CLI example.
Requests served entirely by a page cache or CDN without loading the plugin
cannot be captured; use server/CDN log ingestion for those visits.
The queue requires a writable local filesystem with working flock support.

1.0.2

  • Use response Content-Type at shutdown instead of request Accept to capture crawler page visits.
  • Wait for concurrent queue mutations using a filesystem lock instead of dropping events after option-lock retries.
  • Clarify cron scheduling and cache coverage requirements.

1.0.1

  • Warn administrators when DISABLE_WP_CRON is enabled and queued events require an external cron runner.
  • Track .xml and .txt visits, including requests with non-HTML Accept headers.
Back to top