Comments on: HTTP Caching http://www.subbu.org/blog/2005/01/http-caching HTTP, REST and some Cycling Thu, 20 Nov 2008 14:26:31 +0000 http://wordpress.org/?v=2.7-beta3 hourly 1 By: Subbu Allamaraju http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-27 Subbu Allamaraju Thu, 13 Mar 2008 15:50:25 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-27 Try setting the Last-Modified and Cache-Control headers. For example: Last-Modified: Mon, 18 Feb 2008 17:57:32 GMT Cache-Control: public, max-age=864000 Please be aware that "public" means that any intermediate cache can store the response. Set this to "private" if the response is user-specific and should not be cached in public caches. Try setting the Last-Modified and Cache-Control headers. For example:

Last-Modified: Mon, 18 Feb 2008 17:57:32 GMT
Cache-Control: public, max-age=864000

Please be aware that “public” means that any intermediate cache can store the response. Set this to “private” if the response is user-specific and should not be cached in public caches.

]]> By: R.Prakash http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-26 R.Prakash Thu, 13 Mar 2008 03:40:17 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-26 Hi friends, How to cache a web page in the client machine's tepmorary Internet files folder using the struts app. Hi friends,
How to cache a web page in the client machine’s tepmorary Internet files folder using the struts app.

]]> By: Subbu Allamaraju http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-25 Subbu Allamaraju Thu, 10 Jan 2008 11:47:51 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-25 Ron, >> If Knuth is right that premature optimization is the root of all evil, then it's a Good Thing that there's more information on preventing browsers from caching than on improving cacheability. The user experience penalty for failing to cache is merely slowness (possibly imperceptible). The user experience penalty for caching incorrectly is getting the wrong data. You are right about user experience. However, I would argue that not setting cache headers correctly would provide a deteriorated user experience. Being universal HTTP-clients, browsers do expect certain clues about how to treat a given markup, and caching headers do provide one set of clues. >> Given the general bugginess and unreliability of software, I would place worrying about trying to cache dynamically generated product catalog pages very low on the list of things to do (how many minutes, hours, days, or weeks of incorrectness do you want the user to see when you make a product change?) It depends on what kinds of traffic you are expecting and how much resources you have on the server side. As Mark Nottingham once remarked (http://www.mnot.net/blog/2005/11/26/caching), caching is a way of distributing your application. >> The low-hanging cache fruit is in resources that are never fetched directly (from a user perspective), but only via on-page references: images, javascript, css, etc. Ideally, all CMSs would make it easy to generate unique URLs for each revision of each of these resources (and automatically regenerate all resources containing references to them), which could then be marked as "cache forever". This kind of caching is easy to do without fear of ever presenting the user with stale/incorrect data, and it's currently used by very, very few websites in the world. On the contrary, caching of static resources is the first step, and all major web sites spend quiet a bit of effort in this area. In addition to setting long cache expiry times, these sites also offload these resources to CDNs. Subbu Ron,

>> If Knuth is right that premature optimization is the root of all evil, then it’s a Good Thing that there’s more information on preventing browsers from caching than on improving cacheability. The user experience penalty for failing to cache is merely slowness (possibly imperceptible). The user experience penalty for caching incorrectly is getting the wrong data.

You are right about user experience. However, I would argue that not setting cache headers correctly would provide a deteriorated user experience. Being universal HTTP-clients, browsers do expect certain clues about how to treat a given markup, and caching headers do provide one set of clues.

>> Given the general bugginess and unreliability of software, I would place worrying about trying to cache dynamically generated product catalog pages very low on the list of things to do (how many minutes, hours, days, or weeks of incorrectness do you want the user to see when you make a product change?)

It depends on what kinds of traffic you are expecting and how much resources you have on the server side. As Mark Nottingham once remarked (http://www.mnot.net/blog/2005/11/26/caching), caching is a way of distributing your application.

>> The low-hanging cache fruit is in resources that are never fetched directly (from a user perspective), but only via on-page references: images, javascript, css, etc. Ideally, all CMSs would make it easy to generate unique URLs for each revision of each of these resources (and automatically regenerate all resources containing references to them), which could then be marked as “cache forever”. This kind of caching is easy to do without fear of ever presenting the user with stale/incorrect data, and it’s currently used by very, very few websites in the world.

On the contrary, caching of static resources is the first step, and all major web sites spend quiet a bit of effort in this area. In addition to setting long cache expiry times, these sites also offload these resources to CDNs.

Subbu

]]> By: Ron Burk http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-24 Ron Burk Thu, 10 Jan 2008 10:51:54 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-24 <p>If Knuth is right that premature optimization is the root of all evil, then it's a Good Thing that there's more information on preventing browsers from caching than on improving cacheability. The user experience penalty for failing to cache is merely slowness (possibly imperceptible). The user experience penalty for caching incorrectly is getting the wrong data.</p> <p> Given the general bugginess and unreliability of software, I would place worrying about trying to cache dynamically generated product catalog pages very low on the list of things to do (how many minutes, hours, days, or weeks of incorrectness do you want the user to see when you make a product change?) </p> <p> The low-hanging cache fruit is in resources that are never fetched directly (from a user perspective), but only via on-page references: images, javascript, css, etc. Ideally, all CMSs would make it easy to generate unique URLs for each <b>revision</b> of each of these resources (and automatically regenerate all resources containing references to them), which could then be marked as "cache forever". This kind of caching is easy to do without fear of ever presenting the user with stale/incorrect data, and it's currently used by very, very few websites in the world. </p> If Knuth is right that premature optimization is the root of all evil, then it’s a Good Thing that there’s more information on preventing browsers from caching than on improving cacheability. The user experience penalty for failing to cache is merely slowness (possibly imperceptible). The user experience penalty for caching incorrectly is getting the wrong data.

Given the general bugginess and unreliability of software, I would place worrying about trying to cache dynamically generated product catalog pages very low on the list of things to do (how many minutes, hours, days, or weeks of incorrectness do you want the user to see when you make a product change?)

The low-hanging cache fruit is in resources that are never fetched directly (from a user perspective), but only via on-page references: images, javascript, css, etc.
Ideally, all CMSs would make it easy to generate unique URLs for each revision of each of these resources (and automatically regenerate all resources containing references to them), which could then be marked as “cache forever”.
This kind of caching is easy to do without fear of ever presenting the user with stale/incorrect data, and it’s currently used by very, very few websites in the world.

]]> By: Subbu Allamaraju http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-23 Subbu Allamaraju Sat, 13 Jan 2007 19:59:36 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-23 I've not looked at what Tomcat does, but I would assume that it would set a Last-Modified header based on the file's modified date. I’ve not looked at what Tomcat does, but I would assume that it would set a Last-Modified header based on the file’s modified date.

]]> By: Bryce http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-22 Bryce Fri, 12 Jan 2007 22:54:12 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-22 How about static images that Tomcat or other containers serve. How you specificy an Expires or Last-Modified for these? How about static images that Tomcat or other containers serve. How you specificy an Expires or Last-Modified for these?

]]> By: Ravi http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-21 Ravi Fri, 01 Sep 2006 03:56:02 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-21 I need to get rid of HTTP 304 msg's from the network traffic. I am using Fiddler - HTTP Debugging proxy to findout the trafic. We have around 40% traffic with 304 status code. I used HTTP header "Expires" and also "Last-Modified", but it was not helpful for me. We have lot of static elements such as Image/gif's, .js files, CSS. We are using jakarta-tomcat-4.1.31. Can you help me out. Is there are way I can directly set(Caching Static Elements) in tomcat server rather than changing the code. If not can you please let me know how to fix this in my code. I need to get rid of HTTP 304 msg’s from the network traffic. I am using Fiddler - HTTP Debugging proxy to findout the trafic. We have around 40% traffic with 304 status code. I used HTTP header “Expires” and also “Last-Modified”, but it was not helpful for me. We have lot of static elements such as Image/gif’s, .js files, CSS. We are using jakarta-tomcat-4.1.31. Can you help me out. Is there are way I can directly set(Caching Static Elements) in tomcat server rather than changing the code. If not can you please let me know how to fix this in my code.

]]> By: pratik http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-20 pratik Thu, 24 Aug 2006 06:59:24 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-20 i read all the help from your site.but i found some problem in my asp program. i firefox backbutton not working when we use fckeditor(editor) so please help me........ i read all the help from your site.but i found some problem in my asp program.
i firefox backbutton not working when we use fckeditor(editor)
so please help me……..

]]> By: Datta http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-19 Datta Thu, 24 Aug 2006 03:53:51 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-19 Hi Nice blog. Really helpful and inspiring. I have a question can u help me. In my struts applicatyion i have to implement a search feature which searches files in different hosts(different machines) how to implemet this feature. Plz suggest me if there are any plug-ins/webservices available. Also kindly let me the know the development procedure. Thanks a lot Hi
Nice blog. Really helpful and inspiring. I have a question can u help me.
In my struts applicatyion i have to implement a search feature which searches files in different hosts(different machines) how to implemet this feature. Plz suggest me if there are any plug-ins/webservices available. Also kindly let me the know the development procedure.
Thanks a lot

]]> By: Alex Popescu http://www.subbu.org/blog/2005/01/http-caching/comment-page-1#comment-18 Alex Popescu Tue, 27 Jun 2006 11:57:24 +0000 http://wp.subbu.org/2005/01/http-caching/#comment-18 Hi! Nice entry. I am wondering if and if yes what is the influence of browser settings (IE: Check for newer versions of stored pages, FF: browser.cache.check_doc_frequency) and the hints you posted. ./alex -- .w( the_mindstorm )p. --- (http://themindstorms.blogspot.com) Hi!

Nice entry. I am wondering if and if yes what is the influence of browser settings (IE: Check for newer versions of stored pages, FF: browser.cache.check_doc_frequency) and the hints you posted.

./alex
–
.w( the_mindstorm )p.
—
(http://themindstorms.blogspot.com)

]]>