Paragraph Health is an administration toolkit that helps site builders and developers monitor, audit and clean up the data generated by the Paragraphs module.
When a site relies heavily on Paragraphs, the underlying database tables (paragraphs_item, paragraphs_item_revision, paragraph__*, paragraph_revision__*) can grow very fast, accumulate orphan rows and leave behind obsolete revision tables. Paragraph Health surfaces all of that in a single administration UI and provides safe cleanup tools — both through the web interface and through Drush.
Key features
- Storage overview report. A dashboard under Reports → Paragraph Health that shows the total database size, how much of it is used by paragraphs, the count of paragraph items and revisions, and a per-type breakdown of every paragraph field table with row counts and size.
- Historical charts. A cron collector snapshots the size of each paragraph table over time and renders growth charts (Chart.js) so you can spot tables that are exploding.
- Orphan paragraphs detection. Lists paragraphs whose
parent_id/parent_typeis missing or points to a node that no longer exists. Includes filtering by bundle, pagination, single deletion and batch “delete all” with progress feedback. - Orphan revision tables cleanup. Detects unused
paragraph_r__HASHtables that were left behind by removed fields or bundles, shows their size and row count, and lets you drop them safely after confirmation. - InnoDB table optimizer. Batch UI to run
ALTER TABLE … ENGINE = InnoDBon selected paragraph tables to reclaim space. - Paragraph Type usage report. Shows where each paragraph type is being used across content.
- Field Explorer. Given a node and one of its revisions, renders the value of every paragraph field of that revision — useful to audit revision content.
- Queue processor. Web UI + Drush command to process the
entity_reference_revisions_orphan_purgerqueue with progress feedback, which is required before dropping orphan revision tables. - CSV export of the storage report.
- Security-focused cleanup. All destructive SQL is guarded by a strict whitelist of paragraph-related table names and audit-logged with the acting user ID.
Drush commands
drush paragraphs_health:drop-r-tables— scan and drop orphanparagraph_r__*tables.drush paragraphs-health:delete-paragraphs-orphans— delete every orphan paragraph, optionally filtered by--type.drush paragraphs-health:delete-paragraphs-orphans-random— delete a random subset, useful for large sites.drush paragraphs-health:paragraphs-orphan-count— count orphan paragraphs by type.drush paragraphs-health-queue-processor:run <em>queue_name</em>— process a queue with progress output.drush paragraphs_health:queue-size— report queue table size and per-queue item counts.
When to use it
- Your database is growing much faster than the content on the site would suggest.
- You suspect there are orphan paragraphs from deleted nodes, failed migrations, or aborted content operations.
- You’ve changed or removed paragraph fields/bundles and want to remove the leftover revision tables.
- You need visibility into which paragraph types and fields are the heaviest in storage.
Requirements
- Drupal 10 or 11.
- Paragraphs module.
- Database: MySQL / MariaDB is fully supported. PostgreSQL is partially supported for size reports; SQLite has limited support.
Permissions
All administration screens and Drush commands are guarded by the administer site configuration permission. Always run a database backup before executing destructive operations.
Project information
- Created by cristianodemari on , updated
Stable releases for this project are covered by the security advisory policy.
Look for the shield icon below.
