What Is a Server?

The word covers two things at once. MDN distinguishes the hardware, a computer connected to the internet that stores the site's files, such as HTML, images, stylesheets and JavaScript, from the software, which at a minimum is an HTTP server: the program that understands web addresses and the HTTP protocol and delivers the content.

The server is the other half of every visit. The browser is the client that asks for a page, and the server holds it and sends it. One computer can host many sites, and a large site can be spread across many computers.

When a browser needs a file, it requests it over HTTP. The server receives the request, finds the document and sends it back with a status code, such as 200 if everything went well or 404 if the page does not exist.

MDN distinguishes two kinds of web server. A static one sends files exactly as they are stored. A dynamic one adds an application server and usually a database, and builds the page on each request, for example by filling an HTML template with content from the database. A site with thousands of pages can run on a few templates and a large database this way.

There are often other servers between the user and the origin server. According to MDN, a CDN is a group of servers spread over many locations that store copies of the data and answer from the one closest to the user. Requests the CDN answers never reach the origin server.

Server speed is the first thing that counts in a visit. TTFB, time to first byte, measures how long it takes for the start of the response to arrive. web.dev gives a rough guide: 0.8 seconds or less is good, and more than 1.8 seconds is poor. TTFB is not one of the Core Web Vitals, but it comes before them: if the server is slow, everything else starts late.

The server also determines how much Google crawls. According to Google, its crawlers calculate a crawl capacity limit so as not to overload the server. If the site responds consistently and its response times stay stable or improve, the limit goes up. If it slows down or responds with 5xx errors or a 429, the limit goes down and Google crawls less.

We measured this for zds.es on 15 October 2026, with ten requests to a glossary page and ten to a JavaScript file from the template. Cloudflare marks the HTML pages as DYNAMIC, so the origin server answers them: the first byte arrived after 0.17 seconds on average. The JavaScript file comes from Cloudflare's cache and arrived after 0.06 seconds. Both values are well below the 0.8-second guide.

• Measure the TTFB of your important pages, not just the homepage.
• Serve static files through a CDN or a cache, so the origin server only handles what changes.
• Check the Crawl Stats report in Search Console, which shows the average response time and the host status.
• Choose hosting with enough capacity for peaks in visitors and crawling.
• Keep the server software up to date.

• Measuring speed only with a full browser cache and missing how long the server takes.
• Confusing the origin server with the CDN and looking for the problem in the wrong place.
• Letting the server return 5xx errors for days without noticing.
• Locking out crawlers with security rules that are too strict.
• Assuming good hosting makes up for a very heavy page.