Availability for public-sector sites
This folder is for IT teams, architects and service owners in public bodies who must keep a site reachable: during peaks, through planned maintenance and when a machine fails. It covers what one server gives you, what several web servers require, database availability, the caches in front of PHP, maintenance windows, sessions, capacity planning and how to set availability targets. The detailed server procedures are in the hosting section for hosting companies, especially Several web nodes and Clustering and shared storage; these pages add the public-sector view: continuity duties, central network teams, change calendars and decisions owned by the organization.
Pages
| Page | Answers |
|---|---|
| High availability overview | What one server gives, what several require, and which single points of failure remain. |
| Several web servers behind a load balancer | What is shared, trusted proxies, cron on one node, deploys and clears on every node. |
| Shared storage and the DFS cluster | How several servers share uploaded files, and what the storage and database teams provide. |
| Database high availability | What Exponential does for database availability, and how to switch to a standby. |
| Read replicas | How MySQL and MariaDB sites send reads to replicas, and when that helps. |
| The static cache | Why it ships off, and the rare cases where it still fits. |
| Velocity response cache and HTTP cache | The caches that answer before rendering, their freshness and how to clear them. |
| Cache warming after changes | How to warm the site after a deploy, a restore or before an announcement. |
| Maintenance mode and planned downtime | What visitors and monitors see, who gets in, and planning a window. |
| Sessions across servers | Keeping editors signed in across servers, and what a failed node means. |
| Capacity planning | The figures the repository gives and how to measure your own site. |
| Availability targets and service levels | Which commitments you can define, what they depend on and how to measure them. |
Reading order
- Decide what you need: Availability targets and service levels, then High availability overview.
- Keep one server fast and predictable: the caches, cache warming, capacity planning and maintenance mode.
- Only when the targets require it, several servers: load balancer, shared storage, sessions, database availability and read replicas.
- Read The static cache only if someone asks for it.
Continue with Backups and recovery; hosting choices are in Hosting.
Related pages
- Up: Operations
- Start of the section: Governments: start here
- Free documentation: this page is free to copy, translate and improve under the GNU FDL 1.3.
Become an Exponential Partner
Partners share their information, provide funding, contribute back and develop the products with us, and that is how Exponential grows. A partner website and dashboard are planned for mid-2027: please become a partner today to speed up our growth rate and land new features faster.