When a cloud-hosted website slows down, first identify which resource is under pressure. More disk space will not fix a CPU bottleneck, and a fast page served from cache can hide a slow checkout or search request. Use account metrics alongside application logs and real customer symptoms.
IN THIS ARTICLE
01 Collect a useful baseline
Record the time, affected URL, visitor activity and any recent deployment. Open the resource information available in your hosting workspace or cPanel. View a period that includes the incident rather than a single current reading. Plan limits and the exact metrics shown depend on the hosting configuration.
- CPU: work performed by application requests, scheduled tasks and background processing.
- Memory: working space used by running processes. Repeated failures can occur even when storage remains available.
- I/O and IOPS: the rate and number of storage operations. Large backups, logs and repeated file scans can affect these.
- Concurrent processes or requests: the number of operations active at the same time. Slow requests can occupy capacity for longer.
- Disk space and file count: stored data and number of files, including mail, caches and backups.
02 Reduce avoidable pressure
- Check scheduled tasks for overlap. A backup or import should not restart every few minutes while its previous run is still active.
- Review application errors and slow operations. Repair loops, failed external calls and inefficient database queries before adding capacity.
- Use caching suited to the application. Keep carts, checkout, logged-in sessions and personalised pages out of inappropriate public page caches.
- Resize large media and rotate logs using supported controls. Download a verified backup before deleting anything, and never remove database or application files simply because they are large.
- Repeat the same test after one change and compare the same metrics and customer journey.
When should I upgrade my Cloud Hosting resources?
If normal, legitimate usage repeatedly reaches the current allocation, send support the metrics, incident times and expected growth. Review available upgrade options, recurring price and any migration requirement. Do not assume a plan expands automatically or that an upgrade will take effect without an order.
A website that has spare resources but still responds slowly may be waiting on an external API, a remote database or frontend scripts. Share that evidence so the next step addresses the actual cause.