Hardening and speeding up Inertia's high-traffic WordPress publication
We stabilized Inertia's WordPress publication after repeated traffic-spike outages, then cut load times and put a support plan in place.
Results
- outages during traffic spikes since launch
- 0
- faster server response time
- 71%
- incident response target
- 15 min
The problem
Inertia's site went down whenever a story took off on social media. Pages were rendered uncached for logged-out readers, a plugin conflict was leaking database connections, and there was no monitoring, so the team learned about outages from readers.
Our approach
-
Audited the stack end to end, including plugins, queries, caching, and hosting configuration, and ranked fixes by impact and risk.
-
Removed or replaced eleven plugins, rewrote the three slowest queries, and added a persistent object cache.
-
Put full-page edge caching in front of WordPress with purge rules tied to publishing events.
-
Hardened the install with a web application firewall, least-privilege roles, two-factor sign-in, and automated off-site backups.
-
Set up uptime, error, and Core Web Vitals monitoring with alerts routed to a Seahawk on-call engineer.
Project story
The problem with success
For a publisher, a story that spreads is the best possible outcome. For Inertia it meant a crashed site, because every page view hit PHP and the database directly. The team had added plugins over the years to fix individual problems, and those plugins were now part of the problem.
Audit, then fix in order
We started with a written audit that ranked every issue by impact and risk. The first fixes were the safe, high-impact ones: edge caching for logged-out readers and a persistent object cache. Then we worked through plugin conflicts and slow queries, testing each change in staging under simulated load.
Security as a baseline
The audit also found weak spots: shared admin accounts, no firewall, and backups stored on the same server as the site. We moved to individual accounts with two-factor sign-in, added a web application firewall, and set up encrypted off-site backups with tested restores.
Ongoing support
Inertia now has monitoring that alerts a Seahawk engineer before readers notice a problem, plus a monthly maintenance window for updates and performance checks.
“The site used to fall over on our best days. Now a story can go viral and nobody on the team has to think about the servers.”