Apache 2.2.15 on RHELS 6.1
Using mod_pagespeed on a server behind https (implemented by the network's Reverse Proxy).
All html urls are written as "//server.example.com/path/to/file.css" (so, without the protocol specified).
Problem : using the default configuration, pagespeed rewrites the urls as "http://server.example.com/path/to/file.css"
I'm trying to figure out how to have it rewrite the urls as https (or leave it unspecified as //).
After reading the documentation, I tried using ModPagespeedMapOriginDomain like this
ModPagespeedMapOriginDomain http://localhost https://server.example.com
Also tried
ModPagespeedMapOriginDomain http://localhost //server.example.com
ModPagespeedMapOriginDomain localhost server.example.com
... To no avail. Urls keep being rewritten with "http://".
Question: how can I have pagespeed use https instead of http in its urls?
Full pagespeed config here, if needed
It turns out mod_pagespeed does not work with "protocol-relative" urls.
Still, the issue is bypassed if you enable trim_urls
ModPagespeedEnableFilters trim_urls
Be mindful of the potential risks (depending on your javascript codebase, ajax calls could break or produce unexpected html).
Adding this to your configuration might work:
ModPagespeedRespectXForwardedProto on
That works, if your reverse proxy forwards the X-Forwarded-Proto header in its requests.
That request header tells PageSpeed what the original protocol was that was used for the request at the loadbalancer, and thereby hands it all it needs to know to correctly rewrite urls.
Related
I have an autoscaling application, which uses an ELB to terminate SSL, and forward 443 requests to http 80 on the instances.
This works fine, except that if I do a redirect within apache, it implicitly uses http rather than https.
for example, I have the following rewrite rule...
RewriteRule ^/v10/(.*) /v11/$1 [R]
If I do the following...
curl -v "https://[mydomain.com]/v10/somescript.js"
I get a 302 response like this...
Location: http://[mydomain.com]/v11/somescript.js
... which then breaks my application, as the browser won't load the script via http.
Of course this makes sense, since as far as apache is concerned, it is http. However aside from re-writing all my redirect rules, I'm wondering if there is some way to convince apache to implicitly generate https redirects, even though it isn't terminating the SSL?
Currently we are using the default wirecloud template. But sinde we enabled SSL and redirect every request to the ssl port I would love to change the urls of static ressources to start with https to avoid mixed content warnings.
Is there a simple way to change the urls to always start wit hhttps instead of http?
That's done automatically, except if WireCloud is behind a proxy (so requests comes using HTTP instead of HTTPS). In those cases you can force WireCloud to use https links by adding this line into the settings.py file:
FORCE_PROTO = "https"
See this link for more info.
I'm a bit out of my depth here and nothing I have found quite addresses my problem. Si any and all suggestions are most welcome.
I've got tomcat6 running on CentOS 6.5 hidden behind an apache server (v2.2.15) and I am using Apache's mod_proxy to expose the tomcat webapps, which are running on port 8080. The tomcat hosts one production application and several development applications. On the apache side, both a Drupal site and the aforementioned tomcat production application are on the same domain and, thanks to rewrite rules, all requests to this domain are changed to https. The development sites are reached via subdomains and do not get re-written as https requests.
For the most part, this arrangement works fine. But parts of the tomcat apps are AJAX (calling a Java Struts 1.2 backend). Most of those requests are handled OK. But a few AJAX requests result in redirects (i.e., forward.setRedirect(true)) and that redirect is http (I guess because the container itself is not secure). As a result, I run into cross site scripting issues. I imagine I can use CORS headers to avoid the problem. But that seems like a hack. Is there a relatively painless way I can use to have tomcat send redirects back as https without making tomcat handle ssl directly?
Cris
You could configure the RemoteIpValve in Tomcat:
Another feature of this valve is to replace the apparent scheme
(http/https) and server port with the scheme presented by a proxy or a
load balancer via a request header (e.g. "X-Forwarded-Proto").
To configure Apache to forward the original protocol in the X-Forwarded-Proto header, add a RequestHeader directive in your Apache config, e.g.:
<VirtualHost *:443>
RequestHeader set X-Forwarded-Proto "https"
...
Note that in Tomcat 7, there is also a RemoteIpFilter.
You don't need to do anything special. It already works. Make sure you set the "redirectPort" in server.xml to Apache's HTTPS port, usually 443, and add the following to your <security-constraint> sections for resources you want secured by HTTPS:
<user-data-constraint>
<description>HTTPS</description>
<transport-guarantee>CONFIDENTIAL</transport-guarantee>
</user-data-constraint>
Late to the game here but others may find this-- we had a similar setup and issue where everything worked fine until the application started using ajax posts which did redirects for the response. The fix was to use mod_header in apache to rewrite redirects using "Header edit Location"
http://httpd.apache.org/docs/current/mod/mod_headers.html
Header edit Location ^http://www.example.com/ https://www.example.com/
This went unnoticed prior to the ajax redirects because the browser has no problem doing page level redirects to http (which apache would then redirect back to https). But the ajax cross-site prevention halts at the initial http missing out on that would then be redirected to https by a subsequent request.
I have Apache installed on my server and I need to redirect from http to https. The reason for this is our load balancer solution cannot hand https so requests come in on http and then we transfer them to https using the below lines in the httpd.conf file.
<VirtualHost 10.1.2.91:80>
Redirect 302 /GladQE/link https://glad-test.com/GladQE/link.do
</VirtualHost>
This works fine for GET requests but POST requests will lose the parameters passed on the URL. What would be the easiest way to perform this redirect and maintain POST params?
I need to get from http://glad-test.com/GladQE/link.do to here https://glad-test.com/GladQE/link.do maintaining POST params
Thanks
Tom
You can try with the HTTP status code 307, a RFC compilant browser should repeat the post request.
Reference: http://en.wikipedia.org/wiki/List_of_HTTP_status_codes
In contrast to how 302 was historically implemented, the request
method is not allowed to be changed when reissuing the original
request. For instance, a POST request should be repeated using another
POST request.
To change from 302 to 307, do that:
<VirtualHost 10.1.2.91:80>
Redirect 307 /GladQE/link https://glad-test.com/GladQE/link.do
</VirtualHost>
Standard Apache redirects will not be able to handle POST data as they work on the URL level. POST data is passed in the body of the request, which gets dropped if you do a standard redirect.
You have an option of either using a PHP script to transparently forward the POST request, or using a combination of Rewrite (mod_rewrite) and Proxy (mod_proxy) modules for Apache like follows:
RewriteEngine On
RewriteRule /proxy/(.*)$ http://www.example.com/$1 [P,L]
P flag passes the request to the Proxy module, so anything that comes to your site (via GET or POST doesn't matter) with a URL path starting with a /proxy/ will transparently be handled as a proxy redirect to http://www.example.com/.
For the reference:
http://httpd.apache.org/docs/current/mod/mod_rewrite.html
http://httpd.apache.org/docs/current/mod/mod_proxy.html
Either your public facing website MUST use SSL to protect confidentiality or there is no sensitive data enver passing through it, and no possibility that your site will ever be used for a lauinchboard for sslstripping (there's a very good reason why Google serve up search results over HTTPS).
If you are not encrypting traffic between browser and your site then why are you trying to encrypt them between your load balancer and your webserver? If you do happen to have a SSL termination outside the load balancer (a very silly approach) then using HTTPS between the load balancer and the webserver is far from efficient. The question also implies lots of other security problems like session fixation/sniffing and SSLStripping vulnerabilities.
We have website e.g. http://www.acb.com which points to a hardware load-balancer which is suppose to load-balance two dedicated server. Each server is running apache as a frontend and uses mod_proxy to forward request to tomcat.
Some pages of our website require SSL like https://www.abc.com/login or https://www.abc.com/checkout
SSL is terminated at hardware load-balancer.
When I configured mod_pagespeed it compressed, minimized and merged css file and rewrote them with an absolute url http://www.abc.com/css/merged.pagespeedxxx.css instead of relative url /css/merged.pagespeedxxx.css.
It works fine for non ssl pages but when I navigate to an ssl page such as https://www.abc.com/login all the css and js files are blocked by browser like chrome as their absolute url is not using ssl.
How can I resolve this issue ?
Check for https string in this documentation and this one.
You should show us in your question your current ModPagespeedMapOriginDomain && ModPagespeedDomain settings.
From what I understand from these lines:
The origin_specified_in_html can specify https but the origin_to_fetch_from can only specify http, e.g.
ModPagespeedMapOriginDomain http://localhost https://www.example.com
This directive lets the server accept https requests for www.example.com without requiring a SSL certificate to fetch resources - in fact, this is the only way mod_pagespeed can service https requests as currently it cannot use https to fetch resources. For example, given the above mapping, and assuming Apache is configured for https support, mod_pagespeed will fetch and optimize resources accessed using https://www.example.com, fetching the resources from http://localhost, which can be the same Apache process or a different server process.
And these ones:
mod_pagespeed offers limited support for sites that serve content through https. There are two mechanisms through which mod_pagespeed can be configured to serve https requests:
Use ModPagespeedMapOriginDomain to map the https domain to an http domain.
Use ModPagespeedLoadFromFile to map a locally available directory to the https domain.
The solution would be something like that (or the one with ModPagespeedLoadFromFile)
ModPagespeedMapOriginDomain http://localhost https://www.example.com
BUT, the real problem for you is that apache does not directly receive the HTTPS requests as the hardware load balancer handle it on his own. So the mod-pagespeed output filter does not even know it was requested for an SSL domain. And when it modify the HTML content, applying domain rewrite maybe, it cannot handle the https case.
So... one solution (untested) would be using another virtualhost on the apache server, still HTTP if you want, dedicated to https handling. All https related urls (/login,/checkout,...) would then be redirected to this specific domain name by the hardware load balancer. Let's say http://secure.acb.com. This name is only in use between the load balancer and front apaches (and quite certainly apache should restrict access to this VH to the load balancer only).
Then in these http://secure.acb.com virtualhosts mod_pagespeed would be configured to externally rewrite domains to https://www.example.com. Something like:
ModPagespeedMapOriginDomain http://secure.example.com https://www.example.com
Finally the end user request is https://www.example.com/login, the load balancer manages HTTPS, talk to apache with http://secure.example.com, and page results contains only references to https://www.example.com/* assets. Now when theses assets are requested with an https domain request you still have the problem of serving theses assets. So the hardware load balancer should allow all theses assets url in the https domain and send them to the http://secure.abc.com virtualhosts (or any other static VH).
This sounds like you configured the rewritten URL as http://www.abc.com/css/merged.pagespeedxxx.css yourself - therefor: Try to use a protocol-relative URL, e.g. remove http: and just state //www.abc.com/css/merged.pagespeedxxx.css - this will use the same protocol as the embedding page was requested in.
One of the well standardized but relatively unknown features of URLs