Can't propagate Cache-Control headers with Surrogate-Control header

My application may set the following response header to cause webcache to process the esi:include's:
Surrogate-Control: content=ORAESI/9.0.4, max-age=3600
It and also may set the following intended for the browser cache:
Cache-Control: private
or say:
Cache-Control: max-age=3600
However Webcache removes this and always adds the following whenever surrogate-control has been set:
Cache-Control: max-age=0
This means I can't have browser caching and esi page compilation, just one or the other.
This seems to be designed behaviour can someone explain why this is and if it can be worked around?
I realise that the Cache-Control header should be ignored by webcache but why cant I propagate it to higher caches?

Patrik,
You'll need to convert the meta http-equiv tags into actual HTTP headers before sending for Web Cache to be able to parse it.

Similar Messages

Cache-control header

I have an authenticated webapp with a servlet that send information back to the client browser that is not particularly sensitive. As such, I'd like for the client to be able to cache the retrieved data for a reasonable amount of time.
Thus, I've added the following code to my servlet:
long expireTime = System.currentTimeMillis() + (7 * 24 * 3600 * 1000); // Cache for one week
response.setDateHeader("Expires", expireTime);
response.setHeader("Pragma", "cache");
response.setHeader("Cache-control", "public");While my expires header is indeed getting set to the date that I specify, AS7 is ignoring my Pragma header and sending two (!) Cache-control headers (as seen in a packet dump using ethereal):
Hypertext Transfer Protocol
    HTTP/1.1 200 OK\r\n
    Server: Sun-Java-System-Application-Server/7 2004Q2\r\n
    Date: Sat, 23 Apr 2005 01:30:15 GMT\r\n
    Content-type: message/rfc822\r\n
    Expires: Sat, 30 Apr 2005 01:30:15 GMT\r\n
    Pragma: No-cache\r\n
    Content-disposition: attachment;filename=msg145206-11461.eml\r\n
    Cache-control: no-cache\r\n
    Cache-control: public\r\n
    Transfer-Encoding: chunked\r\n
    \r\n
    HTTP chunked responseAm I doing something wrong, or is this a bug in AS7?
This is really causing a problem, because I want the browser to toss this data stream over to Outlook Express (the data is an e-mail message) but IE sees the no-cache header and refuses to write the file to disk, and thus, OE has no file to open.
Thanks,
Bill

Oh, BTW we believe this header modification is happening because we are setting a cookie and so WLS sees this and adds its own header. Still, we don't want WLS to do anything with the headers.

Firefox is caching my website page even http cache-control header was in use

Hi
Http header "Cache-Control: private, no-cache, no-store, must-revalidate" was used but firefox still caching my webpage. authenticated page of website can be viewed by using work offline feature of firefox.
Please Help
Thanks
Prasant Sharma

https://developer.mozilla.org/en-US/docs/HTTP_Caching_FAQ
CORS was added to Firefox 29, IIRC. <br />
https://developer.mozilla.org/en-US/docs/Web/HTTP/Access_control_CORS
Have you looked into Page Info, specifically the Meta Tags? <br />https://support.mozilla.org/en-US/kb/page-info-window-view-technical-details-about-page

Reader X doesn't display Pdf stream if response contains header "Cache-Control: no-store, no-cache"

Hi all,
I work on a web application that, among others, generates Pdf documents. It renders them directly within the IE window by "streaming" the content of the Pdf in the response output stream. Note that we also add the header "Cache-Control", "no-store, no-cache, must-revalidate,post-check=0, pre-check=0" to the response.
Everything was fine with previous version of reader but since I installed Adobe reader X the content of the Pdf is not showing any more in my browser.
Here is what I already investigated:
- if I use another machine with an "older" Reader version, it works. If I save the displayed Pdf and try to open it on the machine where X is installed it works
- if I remove the the Cache-Control header, then it works with reader X installation
Do you have any idea what changed between version 9 and X that could lead to this issue ?
To ease diagnostic I created a sandbox environment to reproduce the problem, you can go to the following address to see what's happening (or not in case you have version X installed)
With the Cache-Control header: http://readerxissue.appspot.com
Without the Cache-Control header: http://readerxissue.appspot.com/enableCache.html
I must confess that I am a bit stuck and I wonder if some of you could help.
Thanks a lot
Regards
Vincent

Hello,
We have semiliar problems in sweden with Adobe Reader X 10.1 Swe and Internet Explorer 8.0 with streamed PDF files.
We had some issues and got them resolved by the following
Upgraded to Adobe Reader 10.1.2.45 Eng
- Print Issue:http://helpx.adobe.com/acrobat/kb/pdf-wont-print-reader-10.html
- Grey box in Internet explorer: http://helpx.adobe.com/acrobat/kb/pdf-opens-grey-screen-browser.html
- Add site as trusted: Edit-> Preferences, unbox Enable Enhanced Security + add the host/url to site that is whitelisted to send pdf files
(the Automaticly trust sites from mu Win OS security zones doesnt work for us)
The thing is we run MUI pack on our Citrix servers and want Adobe Reader in Swe but it havent been translated yet...
So we have to wait for the swedish release on Adobe Reader X 10.1.2
Thanks,
Tony Van Der Haagen
IT-Mästaren
Sweden

Cache control directives in iPlanet 4.1 (SunOS 5.7)

Using the admin gui (Content Mgmt->Cache Control Directives), I set the cache control to "No Cache" for the entire server; however, clients still seem to cache pages from the server (IE6 on winxp does not go to the server for pages it has seen before). I looked at the response headers using wget on solaris (as well as wfetch on winxp), and I don't see the Cache-Control header. Here is what I get back after requesting a page from the server:
HTTP/1.1 200 OK
Server: Netscape-Enterprise/4.1
Date: Wed, 31 Dec 2003 14:01:13 GMT
Content-type: text/html
Content-length: 120
Connection: keep-alive
I restarted the server, and the admin gui shows the correct ("No Cache") setting for the server.
Thanks in advance for any ideas on how to get the server to ask the client not to cache the pages.

Can you show us your obj.conf file?
Thanks
Manish

Cache-control on IPlanet 6.0 SP5 doesn't work

Hi.
I am desperately trying to set cache-control max-age=0/no-cache to every file that is sent by the IPlanet from a special directory. I used the web console to set /dir/* to no-cache or max-age=0 but althoug the settings are applied and the server is restarted the HTTP header is still not touched in any way.
I read to changes of SP6 and SP7 but couldn't find anything regarding this problem so I haven't upgraded yet.
obj.conf says:
<Object ppath="/var/apps/weblogic/myserver/public_html/mydir/*">
PathCheck fn="set-cache-control" control="no-cache"
</Object>
This is placed directly after the default object.
Any help is highly appreciated. Thanks a lot,
Thomas

There's a big difference between the set-cache-control SAF and the HttpServletResponse.addHeader API. set-cache-control understands the HTTP protocol version in use and, if that version supports Cache-control header fields, adds a Cache-control header. HttpServletResponse.addResponse blindly adds any header you specify.
If you want to use HttpServletResponse.addHeader to add a Cache-control header to a response in Web Server, you can do so. The meaning of that header isn't defined by the HTTP/1.0 protocol, however.

Toplink cache coordination problem with opmn lookup

Dear all,
We encounter some problems when we use opmn url and jms to realize the toplink cache coordination.
Scenario:
1. Using oracle application server 10g, create three processes(jvm) on one oc4j instance
2. Using oc4j in memory jms server for cache coordination
3. Using opmn url to lookup jms topic connection factory and connection, opmn url: opmn:ormi://shasudv4:6004:OC4J_TTS/tts
When we start oc4j instance, we find the log, all the toplink cache coordination properites has been set to toplink framework, and remote command manager has been initialized successfully. All in memory jms server has also been started successfully, there are three listeners on the topic.
But, when we test the synchronization, the data can't be synchronized between all the processes, two of them is ok, the rest of them has some problem to coordinate with the other two. There are no exceptions, we have set the level of toplink log to 'all'.
I have checked the thread dump, I find that "HTTPThreadGroup", "RMICallHandler" and "JMSRequestHandler" threads are involved in cache synchronization.
So, we are obliged to change the opmn url to ormi url, so we can only set a bound for rmi ports of processes, because the ormi port is dynamicly assigned.
<port id="rmi" range="12405-12407" />
This seems ok, and involved thread are: "RMICallHander" and "JMSRequestHandler". But we also concern that we have restricted the port bound of ormi port of every instance.
Thanks of any advice.

I raised a TAR with Oracle. The short answer is that you can't do cache coordination either with ADF business objects, or with Toplink Essentials (as opposed to Oracle Toplink, which doesn't yet support JPA).

HTTP Headers - enabling caching and compression with the portal?

Has anyone configured their web server (IIS or Apache) or use a commercial product to flawlessly cache and compress all content generated by the portal?
Compression and caching is critical for making our portal based applictions work for overseas users. It should be doable, just taking advantage of standard HTTP protocols, but implementing this a complex system like the portal is tricky, we seem to be generating different values in the HTTP Headers for the same types of files (such as CSS).
We are running Apache so can't take advantage of the built in compression capabilities of the .net portal. We are running the java vervion. 6.1 mp1, sql server 2000 (portal, search, collab, publisher, studio, analytics, custom .net and java portlets on remote server).
Basically our strategy is to compress all outgoing static and dynamic text content (html, CSS, javascript), and to cache all static files (CSS, javascript, images) for 6 months to a year depending on file type.
Here are some links on the subjects of caching and compression that I have compiled:
Caching & Compression info and tools
http://www.webreference.com/internet/software/servers/http/compression/
http://www.ibm.com/developerworks/web/library/wa-httpcomp/
http://www.mnot.net/cache_docs/
http://www.codeproject.com/aspnet/HttpCompressionQnD.asp?df=100&forumid=322472&exp=0&select=1722189#xx1722189xx
http://en.wikipedia.org/wiki/Http_compression
http://perl.apache.org/docs/tutorials/client/compression/compression.html
https://secure.xcache.com/Page.aspx?c=60&p=590
http://www.codinghorror.com/blog/archives/000807.html
http://www.howtoforge.com/apache2_mod_deflate
http://www.ircache.net/cgi-bin/cacheability.py
http://betterexplained.com/articles/how-to-optimize-your-site-with-http-caching/
http://betterexplained.com/articles/speed-up-your-javascript-load-time/
http://betterexplained.com/articles/speed-up-your-javascript-load-time/
http://www.rubyrobot.org/article/5-tips-for-faster-loading-web-sites
http://betterexplained.com/articles/how-to-optimize-your-site-with-gzip-compression/
http://www.gidnetwork.com/tools/gzip-test.php
http://www.pipeboost.com/
http://www.schroepl.net/cgi-bin/http_trace.pl
http://leknor.com/code/gziped.php?url=http%3A%2F%2Fwww.google.com
http://www.port80software.com/surveys/top1000compression/
http://www.rexswain.com/httpview.html
http://www.15seconds.com/issue/020314.htm
http://www.devwebpro.com/devwebpro-39-20041117DevelopingYourSiteforPerformanceCompressionandOtherServerSideEnhancements.html
http://www.webpronews.com/topnews/2004/11/17/developing-your-site-for-performance-optimal-cache-control
http://www.sitepoint.com/print/effective-website-acceleration
http://nazish.blog.com/1007523/
http://msdn.microsoft.com/library/default.asp?url=/library/en-us/IETechCol/dnwebgen/IE_Fiddler2.asp?frame=true
http://www.fiddlertool.com/fiddler/version.asp
http://www.w3.org/Protocols/rfc2616/rfc2616-sec13.html
http://www.web-caching.com/cacheability.html
http://www.edginet.org/techie/website/http.html
http://www.cmlenz.net/blog/2005/05/on_http_lastmod.html
http://www.websiteoptimization.com/speed/tweak/cache/
http://www.webperformance.org/caching//caching_for_performance.html
http://betterexplained.com/articles/how-to-debug-web-applications-with-firefox/
Edited by tkoenings at 06/18/2007 6:26 AM

Hi Scott,
Does Weblogic platform 8.1 supports netscape? We have developed a portal which
works perfectly on IE but it dies in netscape. Is netUI tags not supported in
Netscape?
Pls reply
manju
Scott Dunbar <[email protected]> wrote:
From a pure HTML perspective Portal does it's rendering with nested
tables.
Netscape 4.x and below have terrible performance with nested tables.
The
problem is not the Portal server but rather Netscape on the client machine.
If IE and/or a recent version of Netscape/Mozilla is not possible then
there are
really only two options:
1) Faster client hardware - not likely to be an acceptable solution.
2) Minimize the number of portlets and the complexity within the portlets.
Neither of these solutions are a great answer, but the 4.7 series of
Netscape is
getting pretty old. Having said that, we've got customers who want to
continue
to use IE 4 :)
Again, though, this problem is, I'm afraid out of our hands. It is the
client
rendering time that is the issue.
cg wrote:
Does anyone know of any known reasons why the 7.0 (did it also with4.0) portal
pages can take up to almost 30 seconds to load in Netscape 4.7? I knowit is a
very generic question but our customer still uses 4.7 and will notuse the portal
b/c it takes so long to load some of the webapps. What the pages willdo when
loading is that the headers will come up and when it gets to the bodyof the page
it seems to stall and then comes up all of a sudden. For some of thepages it
takes 6 seconds and for others it takes about 24-27 seconds.
We have suggested using IE only but that is not an option with allof the customers
and getting a newer version of Netscape is also out of the question.
Any suggestions would be greatly appreciated.--
scott dunbar
bea systems, inc.
boulder, co, usa

Cache-control, Expires headers

I want JNLP files and JAR files to be cached by JNLP clients, but I want to prevent them from being cached in a browser / proxy server / ISP server of some sort (anything that would prevent the JNLP client from getting updates).
To stop a web page being cached I typically use the following:
response.setDateHeader("Expires", 1L);
// Set standard HTTP/1.1 no-cache headers.
response.setHeader(
"Cache-Control", "no-store, no-cache, must-revalidate");
// Set IE extended HTTP/1.1 no-cache headers (use addHeader).
response.addHeader("Cache-Control", "post-check=0, pre-check=0");
// HTTP 1.0 header
response.setHeader("Pragma", "no-cache");This works well in stopping caching of web-pages, and it doesn't stop the Sun Java Web Start client (JRE 1.5) from caching JNLP / JAR files (I don't think it's actually checking any of the above headers).
However, I have read that setting the above headers can cause problems. Does anyone know the official line on this? I couldn't find anything in the JNLP spec.
If it makes a difference, the app will be deployed only on JRE 1.4 or later.

LateNightMovies,
> Hi, thanks for your reply (and sorry for posting my
question
> twice)...
No worries. :)
> ** unfortunately, your solution doesn't work.
It's possible I didn't understand what you're after.
> My objective is to "not have the page stored" (saved) in
the
> cache (it contains data I don't want stored in client
browsers).
Aha. Nine times out of ten, when people say they want to
want to avoid
cacheing, what they mean is, "Get Flash to always ask for new
data," which
is what that blog entry describes. If it's important that the
data not be
cached at all....
> Adding a ? var - does ensure the page is always
retrieved
> from the server - but it still gets saved in the browser
cache.
Right.
> Eg. my url = page.asp?param=# (random number)
>
> The browser cache (Content.IE subdirectories) fill up
with
> files = page[1].html, page[2].html, page[3].html,....
Yup. Well, shoot. This question isn't something I can just
roll off
the top of my head, then. :-/ I should think that your HTML
header would
take care of that.
Are you appending the random number in your LoadVars method
calls? The
HTML page shouldn't be refreshing at all.
I know a few people I can ask for futher input, and I'll
report back to
this thread.
David Stiller
Adobe Community Expert
Dev blog,
http://www.quip.net/blog/
"Luck is the residue of good design."

Controlling Cache-Control headers

We have an application written in BC4J 9.0.3.4 riding atop 9iAS 9.0.3.1.0.
The release notes for this version of 9iAS say
2484551 Microsoft Internet Explorer cannot open some file directories due to HTTP NO-STORE header
If users use Microsoft Internet Explorer 6.0.2600.0000, then they cannot open some attachments, such as Microsoft PowerPoint or voicemail directly. They have to save the file and open it from their file system. The file cannot be opened because Microsoft Internet Explorer 6.0 doesn't save the file to the <code>temp</code> folder when an HTTP no-store header is set.
Saying that this is fixed is, to say the least, iffy. A packet dump clearly shows the following headers being set at some point after JSP filters get a chance to kick in (i.e., probably by mod_oc4j itself):
Pragma: no-cache
Cache-Control: no-store
Surrogate-Control: no-store
Cache-Control: private
There doesn't seem to be any way of turning these headers off other than using .htaccess and mod_headers to explicitly delete them (which is definitely not ideal). Further, my tests indicate that these headers (probably only Cache-Control: no-store) prevent IE5.5 and IE6 both from launching external applications on the basis of Content-Type / extension, and from saving the page at all.
This is, admittedly, a violation of RFC2616 on IE's part: that RFC states that no-store is meant to apply to intermediate proxies, not to user agents themselves. But nonetheless I think Oracle should at least allow pandering to this particular bug; i.e., even if Cache-Control: no-store is on by default, it should be possible to turn it off.
Am I missing something? Is there any way to stop mod_oc4j adding these headers?

I tried the brute-force approach:
Header set Cache-Control: no-cache
Header set Surrogate-Control: no-cache
in .htaccess.
Alas, .htaccess isn't being read, even with
<Directory /oracle/product/ias_903/j2ee/OC4J_DEV/applications>
AllowOverride FileInfo
</Directory>
in the httpd.conf.
This is confusing. Does mod_oc4j somehow ban the processing of .htaccess files even if the AllowOverride says otherwise?

Disable Pagma and Cache-Control headers in SunOne WebServer 6.1

Hi,
I want the [Pragma] and [Cache-Control] headers completely disappear from SunOne WebServer 6.1 JSP responses (like the SunOne WebServer 6.0 SP8, http/1.1 but no [Pragma] and [Cache-Control] headers), can I achieve this?
Thanks,
Harry

Hi Harry
Do you mean that WebServer 6.0 SP8 provide a way to disable these two attributes? If so, that will be a regression issue for WS6..

I there, my question is quite simple, I would like to know if the "apple remote control" can be used with a mac mini and , if so, it can be used to control also the "logic pro" functions (e.g. record, start, stop etc). Thks a lot, Danilo

I there, my question is quite simple, I would like to know if the "apple remote control" can be used with a mac mini and , if so, it can be used to control also the "logic pro" functions (e.g. record, start, stop etc). Thks a lot, Danilo

Good work, thanks for the report.

HT1343 On my 4 1/2 year old Macbook I could select multiple files by highlighting a file and then moving cursor over a file lower in the list, and pressing shift-control-click. I can't do that in now with Mac Pro. What am I doing wrong?

ghlighting a file and then moving cursor over a file lower in the list, and pressing shift-control-click. I can't do that in now with Mac Pro. What am I doing wrong?

Holding down the control key will invoke "right-click."
Depending on the View, holding down Shift will with either toggle the selection (Icon View), or it will make a continuous selection from the previous selection (List View or Column View).
The only way to make a continuous selection in Icon View is to drag out a selection marquee around or through the icons.
I can't remember the behavior of previous OS's.

How can one use Mission Control with two monitors? Please bring Spaces back

How can one use Mission Control with two monitors. With Spaces I could treat each space as a single desktop.
SyBB

I use two monitors at work and have no issue. I have my mail set to use Desktop 1 and iTunes set to the second monitor of Desktop 1. I have browsers set to Desktop 2 and Fusion and RDC set to Desktop 3.
Two things that may help you. In System Preferences > Mission Control, disable the setting "Automatically rearrange spaces...". This screwed with assigning applications to certain spaces. And the other thing is don't use full screen on apps that support it. This just makes your second monitor superfluous.

JRE7 throws exception Can't create cache file but work fine with JRE6

I am using J-free chart for rendering charts. My application is work fine with JRE6 but while using JRE7 i am getting following exception
Caused by: javax.imageio.IIOException: Can't create cache file!
     at javax.imageio.ImageIO.createImageOutputStream(Unknown Source)
     ... 18 more
Caused by: java.io.IOException: The system cannot find the path specified
     at java.io.WinNTFileSystem.createFileExclusively(Native Method)
     at java.io.File.createTempFile(Unknown Source)
     at javax.imageio.stream.FileCacheImageOutputStream.<init>(Unknown Source)
     at com.sun.imageio.spi.OutputStreamImageOutputStreamSpi.createOutputStreamInstance(Unknown Source)
     ... 19 more

I was not able to reproduce this with Java 7u3 (32bit) on a Windows 7 (32bit).
What specific version of Java were you using at the time of error?
- ShakR

Can't propagate Cache-Control headers with Surrogate-Control header

Similar Messages

Maybe you are looking for