When managing large WordPress multisite installations, backups are crucial for ensuring data safety and disaster recovery. However, not all backup methods are created equal, especially when dealing with hundreds—or even thousands—of subsites under one roof. This is where developers began noticing a troublesome pattern with Jetpack Backup: during large-scale backup operations, site performance would dip or temporarily stall altogether, primarily because the backup process was locking subsite databases. This article explores how the Table Split Technique offered a sophisticated workaround to that issue.
TL;DR
Jetpack Backup’s default behavior of locking entire subsite tables during backups led to significant performance issues on large WordPress multisite networks. Traffic bottlenecks and user-facing errors were commonplace during heavy backup cycles. The implementation of the Table Split Technique allowed for a smarter, asynchronous backup approach that mitigated locking and significantly improved performance. This technique involves splitting table exports, backing up non-essential tables separately, and minimizing locking times.
Understanding the Problem: Database Locking During Backups
Jetpack Backup is widely praised for its ease of use and seamless integration with WordPress, but for large multisite environments, it inadvertently introduced a silent pain point: database locking. When performing full database exports, Jetpack Backup traditionally relies on mysqldump or similar tools, which tend to lock tables, especially when exporting large structures like wp_posts, wp_options, or wp_usermeta.
On usual single-site installations, this interruption might go unnoticed. But on a multisite network running 500+ subsites with their own prefixed tables (e.g., wp_100_posts, wp_101_options), the backup engine attempts to snapshot all of them—causing large lock durations that in turn freeze front-end operations, thwart AJAX calls, and eventually prompt timeouts.
This was especially problematic in these scenarios:
- eCommerce sites unable to process transactions during lock periods
- Content-heavy sites showing stale or missing data to logged-in users
- Admin operations becoming unresponsive—leading to frustrating management experiences
Anatomy of a Multisite Backup Conflict
The crux of the issue lies in how MySQL handles reads and writes. During a logical backup of a table, depending on the transactional engine (e.g., MyISAM or InnoDB), it may either attempt a read lock or even temporarily block writes. Jetpack Backup, attempting to be efficient, pulls in entire tables through reads that scale poorly—especially those with tens of thousands of rows.
For example, a common pattern that users experienced was the following:
- Jetpack initiates a backup at midnight
- On a multisite setup with 700 subsites, each subsite’s tables are exported serially
- During table exports, particularly of dynamic tables (like
commentsorposts), a lock is applied - Concurrent front-end or admin traffic leads to timeouts due to the locked resources
Even with strategies like staggering backups or excluding specific tables, the pinch was still felt. The need for finer granularity became clear: someone needed to stop treating the backup process as a monolith and instead approach it modularly.
The Birth of the Table Split Technique
The solution was inspired by the principle of microservices: breaking large, tightly-coupled operations into discrete, asynchronous ones. Thus was born the Table Split Technique, a method that fundamentally changed how backups were performed within multisites by breaking down table groups logically.
This technique emphasized:
- Classifying tables based on volatility (e.g., logs vs. configurations)
- Segmenting backup jobs to isolate critical and non-critical operations
- Performing parallel exports using lock-free techniques where possible
Instead of backing up the entire subsite database in one go, each subsite’s tables would be divided into:
- Static Tables: Configurations, settings, user roles—rarely change, backed up weekly
- Dynamic Tables: Posts, comments, orders—change frequently, backed up incrementally
- Transient Tables: Cache and session data—often skipped in backups entirely
Implementing the Technique
The first step in implementing this strategy was creating a profiling script that measured query traffic per table across subsites. This identified which tables were both large and frequently accessed. Once tables were categorized, custom backup scripts were built using mysqldump --single-transaction for InnoDB tables, further reducing lock needs.
Additionally, tools like Percona XtraBackup and WP-CLI custom aliases helped pipe table-specific exports into their own job queues. These could then be scheduled staggered through cron jobs or real-time triggers with rate caps—ensuring that even if 10 backups occurred simultaneously, none would focus on the same high-usage table group.
The Outcome: Faster, Lighter, Smarter Backups
After deploying the Table Split Technique across a multisite hosting over 1200 subsites, results were immediate and impactful:
- Frontend response times improved by over 35% during backup windows
- Backup failures dropped by 82% due to fewer timeout-induced errors
- Jobs could be scheduled more predictably, allowing simultaneous read-friendly exports across nodes
Importantly, subsites that once experienced momentary “freezing” during backup hours now ran smoothly – even during sales campaigns or traffic spikes. Site administrators also reported that they could perform dashboard operations hassle-free, no longer at the mercy of unpredictable backup cycles.
Jetpack’s Response and Future Improvements
While the Table Split Technique emerged as a user-originated workaround, its effectiveness sparked discussions around improving the internal operations of tools like Jetpack Backup. Developers proposed new features such as:
- Automatic table grouping and smart scheduling based on change frequency
- Options to integrate with pgBackRest or mysqldumpPlus for fine-grained control
- Real-time backup logs with a lock-impact report
Some of these ideas have already begun to influence plugins outside of Jetpack, ushering in a new generation of WordPress backup tools that prioritize concurrency safety and minimal intrusion.
Conclusion
Jetpack Backup remains a powerful ally in website data protection. But as WordPress installations scale, so do the unintended consequences of monolithic backup strategies. Through innovative practices like the Table Split Technique, performance bottlenecks caused by database locking can be effectively eliminated—paving the way for seamless, high-volume WordPress operations.
For developers and site administrators alike, embracing granular, modular backup processes is no longer a luxury—it’s a necessity for keeping sites resilient, fast, and online during even the heaviest traffic nights. The journey from “half-hour backups and timeout errors” to “transparent, background backups” is not just about tools—it’s about intelligent tooling strategies. The Table Split Technique was one such milestone.
- How Much Do Ultrasound Techs Make in 2025? - November 27, 2025
- Jetpack Backup Locking Subsite Databases During Large Backups and the Table Split Technique That Prevented Conflicts - November 27, 2025
- What Happened to Chatiw? Full Breakdown - November 26, 2025