HTTP request staus fails

I have the same problem. On the same server I have 5.7 installed for a different site with the Status Report module that works flawlessly, no errors or performance issues. When I active Status Report in 6.0 admin pages crawl to a halt. Any ideas, anyone?

Hi ,

I too have the same problem. I tried contacting my webhost, but they need more details regarding php requirements and modules that are being affected by this error.

After searching, I've noticed that no solutions are ever revealed. Is this not a common problem?

I, too have the same problem.

However, I noticed every once in a while it works fine. Not sure why. Seems to make a difference if I go into the Control Panel on my host account and run the "Cron" job manually. After that the Status report is all green!

Could it simply be a time-out error that Drupal is interpreting as a refusal for the HTTP Request?

There is a discussion on this http://drupal.org/node/7881.

Changes need to be made to the /includes/common.inc code to enable http connection through a proxy requiring authentication. There are some examples of what people have done to get around this problem.

I haven't succeeded yet with Drupal 6 as I'm getting syntax errors with my first attempt at modifying common.inc, but the instructions for use aren't clear. I'll let you know if I get it fixed. If anyone else figures it out maybe they can comment here.

Meanwhile there are calls for this proxy authentication issue to be included in core as there are many drupal hosts on intranets behind firewalls and proxy servers.

But at least we know where the solution lies.

I had the same problem but my allow_fopen_url at php.ini was set to "on". I discovered that, if I remove the "offline status" of the site, the "http_requests status fail" message disappears. Tyr it!

Some interesting aspects of this problem have been unearthed - thanks to all who have contributed.

I'm pretty sure that my particular issue is that my web server is behind a firewall which requires proxy authentication.

When I built the web server I was able to configure apt-get for updates using a .conf file:

Acquire::http::Proxy "http://username:password@proxy.myupstreamproxy.net:8080";

and it works like a charm.

What I need to know is how can I do something comparable for my Apache2 PHP5 webserver when it needs to access another website (such as Drupal updates)?

I've scoured Google but have not found anything yet.

Anyone have any ideas?

Thanks

My install is on a fresh RHEL5 box w/ apache2 & Drupal 6.1

My default php.ini already had "allow_url_fopen = On"

No external firewall, but I used system-config-securitylevel to close off all but http & ssh

Since this is a development site I had a password stanza in my .htaccess thus:

AuthType Basic
AuthUserFile /var/www/passwd/htpasswd
AuthName "Cast and Crew Only"
require valid-user

Commenting these lines out fixed the issue and my status report is now all green.

Jake.

... if your drupal installation uses the same hostname as your server.

We have a machine whose name is blah.example.com. We have installed a drupal site for blah.example.com, eg sites/blah.example.com

When system_check_http_request() runs, it tries to fetch http://blah.example.com/admin/reports/request-test and match it to the data received from an output buffer request of the same URL.

This will fail because fsockopen() will resolve blah.example.com to 127.0.0.1 and give you something very different to the output buffer data. Trying to fetch http://127.0.0.1/admin/reports/request-test will result in a 404 because this won't hit drupal.

So drupal_http_request() will think it can't ever access anything, set the drupal_http_request_fails variable to TRUE and never try again.

The reason this is false (in a way) is that although you won't be able to access drupal running on localhost, it will be able to access machines elsewhere.

Using the HTTP proxy patch works (http://drupal.org/node/7881) because the proxy server fetching the url will resolve your machine name (blah.example.com) to its regular IP, and not 127.0.0.1 therefore hitting drupal and getting the data it was after.

As for how to resolve (pun intended) this issue, you can:

a) Not have your drupal install be the same name as your physical machine. Set an alias in your DNS to point to your server, and use that host name as the drupal install (ie, www.blah.example.com)
b) Use the HTTP proxy patch and use a proxy
c) Patch system_check_http_request to use a URL not on your drupal install as a test. It has to be static content and should be short and dependable as always being there. Perhaps drupal.org should host a file like this.

Other than that I don't know.

This obviously isn't the only reason why you would be getting this error. It only applies if your drupal installation uses the exact same name as your server.

I'm having this problem on localhost. I have three feeds set up on Feed Aggregator, and one of them came through one time and the other two never did. I ran cron manually, and I still get the error.

What can I do to fix this on localhost?

Thanks.

I had this problem on Drupal 6.4 hosted by SiteGround. After reading the others here, here is my experience and what fixed it.

1) My cron wasn't really running and I ignored it because my site isn't live yet.
2) While under development, I've got my site password protected using HTTP Auth.
3) Yesterday I started having this error message about the server being unable to access HTTP.
4) I disabled HTTP Auth, ran cron manually, flushed cache, re-enabled HTTP Auth and it works now.
5) I modified the cronjob to use HTTP Auth so it would run :)

I believe because cron didn't run for so long, Drupal locked itself (as someone else said here because of a timeout). Running cron manually with HTTP Auth turned off made it reset itself. It works again too with HTTP Auth re-enabled.

I'm very suddenly getting this one too. Out of nowhere... :(

My apache config blocks access to /admin/ or /user/ from anywhere else except my own static IP address.
Funny thing is this was working, only complained today.
I added the server local IP address to the allowed IPs and the error is gone now.
Thanks to benkant for explaining what happens, so I could find out what I'd done wrong and fix it :)

It happened when I used a fake server name on my destop (via /ets/hosts on the desktop machine).
You just have to make sure that server knows that name too (i.e. add it to the server's /ets/hosts).

I found this sollution on the drupal site (I forgot who originally posted it):

go to: C:\Windows\System32\drivers\etc
and open "hosts"

put an # in front of ::1

This did it for me.
I use Windows Vista.

One other thing to check on Linux.

Make sure bind-utils is installed, it has all the nifty DNS tools - host, dig, nslookup, etc.

I was having issues with the http request status failing as well.

I fixed it by editing the windows/system32/drivers/etc/hosts file and adding "127.0.0.1 localhost" to the bottom of the file (in vista open notepad as administrator).

I suppose you could also just uncomment the same thing towards the top of the file but I wanted to leave the original file as intact as possible.

Hope that helps!

Jeff Swanson
GardenGnomeWorld.com

A quick solution to remove the error message is to edit the following line in system_check_http_request() inside of system.module:

It is trying to do an http request to the drupal instance being run but is failing because it returns a really weird status code... insert a website you can resolve to and it will remove the error ... like so:

This is definitely not the "Drupal way" of fixing the problem, isn't in a patch form, likely isn't attached to the correct issue, and doesn't flow with the Drupal social/political structure but it lets me continue my work ... I would recommend removing your edit to this file when you are done.

updated /etc/hosts on my vmware image (CentOS 5.2) so it knew it's own name and hey presto!!!

I'm running Windows XP Pro and have installed XAMPP, running MySQL 5.1, and Apache. Have just installed Drupal 6.10 and done a few module updates...Can't resolve the status error HTTP request status Failsl.....this has been plagueing me from the beginning. I can't seem to get off the ground here.

Can someone please assist me with suggestions appropriate to running this installation on my local machine as a test environment in XAMPP.

Thanks, WCW

I run Drupal from my home on an internal LAN (non-routable ip's) and was receiving this error. My remedy was to add the external/internal hostnames to the /etc/hosts file: (I used sub domain's for my site)

XXX.XXX.XXX.XXX sub.domain.com
XXX.XXX.XXX.XXX sub2.domain.com

XXX.XXX.XXX.XXX = internal IP address assigned to your web server/virtual host configuration for both domains above.

sub = Is the domain that provides external access to my webserver from the internet ("A" record that points to the WAN side IP of my DSL Modem)

sub2 = Is the domain that provides internal access for my LAN

Hope that helps!

Matt

Hello there!
I solved the issue on Vista by updating C:\Windows\System32\drivers\etc\hosts and C:\Windows\System32\drivers\etc\lmhosts.sam to contain the following mappings:

127.0.0.1 localhost
xxx.xxx.xxx.xxx localhost

where xxx.xxx.xxx.xxx is my static IP.

I also installed the module "Http Request Fail Reset" on drupal (version 6), see http://drupal.org/project/http_request_fail_reset

I hope it helps.

Thanks for all your help with this guys, Id just like to add something that may be overlooked (because I overlooked it!). When adding entries to /etc/hosts to resolve DNS issues, make sure that you include root and sub-domains, example

127.0.0.1 example.com www.example.com

I forgot to do this, and only dawned on me recently. It also explained why this was working intermittently, as sometimes users would access my site via one or the other. In Apache conf, they both resolve to the same vhost, using ServerAlias, but I needed separate entries to help DNS along.

Hope this helps someone else.

Phew. I hope this is the worst Vista has to offer. And I know I am being too optimistic.
Thanks for the fix!

We have an internal development box w/ 1 ip, using internal domain names (site1.localdomain.local, site2.localdomain.local).

Adding:

127.0.0.1 site1.localdomain.local site2.localdomain.local

Fixed the HTTP request fail issue.

Trackback, accidentally showing third party content via ssl can cause this. Like youtube videos.
http://drupal.org/node/227791#comment-2172748