Member Avatar for minimee120

Just would like to know some differences, am learning about it in class

  • 4 Contributors
  • 4 Replies
  • 247 Views
  • 6 Days Discussion Span
  • Latest Post Latest Post by riahc3

Dani AI

Generated

, and covered the basics of "where" a copy lives. The important follow-ups are the HTTP rules that decide who may store a response and for how long. Client (browser) caches are per-user and follow Cache-Control: private or browser heuristics; proxy/CDN caches are shared and obey Cache-Control: public or s-maxage. Freshness is controlled either by expiration (max-age/Expires) or by validation (ETag / Last-Modified + conditional requests).

Common headers and what they mean:

Cache-Control: public, s-maxage=86400        # shared caches may store for 1 day
Cache-Control: private, max-age=3600         # only browser cache, 1 hour
Cache-Control: no-cache                      # must revalidate before use
Cache-Control: no-store                      # do not store at all
Cache-Control: max-age=31536000, immutable   # long-lived fingerprinted assets
ETag: "abc123"
Vary: Accept-Encoding

Validation and cache keys matter. ETag / Last-Modified + If-None-Match / If-Modified-Since let caches return 304 Not Modified instead of full content. The Vary header changes the cache key (for example Vary: Accept-Encoding separates compressed vs uncompressed variants). Responses with Set-Cookie or Authorization are usually treated as private unless explicitly allowed.

Quick diagnostics and practical rules: inspect headers with curl -I or the browser Network tab and look for Cache-Control, Age, Via or X-Cache to detect proxy hits. For production, fingerprint static assets and give them long TTL + immutable; use s-maxage for shared caches; mark personalized or sensitive responses private or no-store. This addresses the traffic/workload point raised by and the "are proxies needed" doubt from : shared caches (including CDNs) still matter at scale and for network-policy use cases, but the right mix depends on traffic patterns and content sensitivity.

Recommended Answers

All 4 Replies

Member Avatar for CimmerianX

It's really just where the copy of the content is stored.

Think of the path taken by your PC to get web content when using a proxy.
1) Your browser makes the request.
2) Browser sends the request to the proxy.
3) Proxy relays the request to public www site.
4) public www site sends content to the Proxy.
5) Proxy sends it to your browser for display.

If you hit a site often, like Daniweb, certain static content doesn't change. Logo Image for example. If the file hasn't changed, there's no need to re-download it. right?

(Basically) Your browser can cache the logo image locally, when the requests go out for the content, there is no need to request the image file again. If you have 50 machines, each one must keep a copy of the logo in local cache after having downloaded it at least once. If the logo changes, 50 machines must pull the file again.

If the proxy is caching, then the logo is stored on the server. The 1st daniweb visitor reguests the page and the images are pulled by the proxy and now stored on the proxy. The next 49 machines that visit daniweb pull the image from the proxy instead of the daniweb host, thus saving the bandwith of pulling the image 49 more times. If the image changes, the server pulls it just once to service the 50 internal machines.

Member Avatar for riahc3

Here is a intresting thought though: In now, a modern area, are proxy servers really neccesary (in the scenario you put of pulling a logo)? With all the bandwidth avaliable to most first class countries....I dont see it logical.

I agree in other scenarios though a proxy server is very important and can save a lot of bandwidth.

Member Avatar for lolafuertes

Is not a bandwidth (speed) problem, is a traffic (volume) one.

Imagine an scenario: 1 web server, 100.000.000 pages accessed each day. If you put some proxies in the middle, all the static parts will flow only from the server once, then from the proxies, while the dynamic part flow always from the server.

This helps to diminish not only the traffic but the server work.

Hope this helps.

Member Avatar for riahc3

Is not a bandwidth (speed) problem, is a traffic (volume) one.

Imagine an scenario: 1 web server, 100.000.000 pages accessed each day. If you put some proxies in the middle, all the static parts will flow only from the server once, then from the proxies, while the dynamic part flow always from the server.

This helps to diminish not only the traffic but the server work.

Hope this helps.

Did not think of the server workload.

Be a part of the DaniWeb community

We're a friendly, industry-focused community of developers, IT pros, digital marketers, and technology enthusiasts meeting, networking, learning, and sharing knowledge.