Case Study
The Server That Fell Over at 3am
Every night at 3:07am, the site went down. Nobody could say why.
Challenge
Frequent unexplained website downtime, lost sales and indexing issues.
Outcome
Identified intensive cron jobs and backup processes competing for resources. Separated time for automated processes and optimised the database.
Key Results
- Reduced server load to maintain stability
- Fine tuned the database for performance
An ecommerce owner was getting 50+ downtime alerts a week. Pages crawled, checkouts timed out, and by morning everything was “fine.”
Their host blamed the site, and the previous developer blamed the host. Meanwhile they upgraded to a bigger server plan twice. The problem didn’t move.
We traced the system calls and found the culprit: a runaway scheduled job fighting a backup process for the same resources, every night. A bigger server just gave it more room to misbehave.
We rescheduled and rewrote the job, tuned the database queries, and added monitoring. Overnight load dropped 80%, and they ended up downgrading their hosting plan.
Throwing money at a performance problem rarely works. Find the cause first.
Site slow or flaky at odd hours? We’ll find out why.
Need help with your website?
Our team repairs, maintains, and rescues websites across Australia.
See our services