Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

ownCloud 9.0.2 security and setup warnings are endlessly reloading #24309

Closed
LukasReschke opened this issue Apr 27, 2016 · 28 comments · Fixed by #25239
Closed

ownCloud 9.0.2 security and setup warnings are endlessly reloading #24309

LukasReschke opened this issue Apr 27, 2016 · 28 comments · Fixed by #25239
Assignees
Milestone

Comments

@LukasReschke
Copy link
Member

Other users have reported this as well in the bug tracker, with ownCloud Proxy I now face this as well. I'm debugging this…

@LukasReschke LukasReschke self-assigned this Apr 27, 2016
@LukasReschke LukasReschke added this to the 9.0.2-current-maintenance milestone Apr 27, 2016
@LukasReschke
Copy link
Member Author

Mixed Content: The page at 'https://█████.owncloudconnect.com/owncloud/settings/admin' was loaded over HTTPS, but requested an insecure XMLHttpRequest endpoint 'http://█████.owncloudconnect.com/owncloud/apps/files/'. This request has been blocked; the content must be served over HTTPS.
Navigated to https://█████.owncloudconnect.com/owncloud/settings/admin

@ghost
Copy link

ghost commented Apr 27, 2016

Dup: #24293 ? :-)

@RobinMcCorkell
Copy link
Member

@LukasReschke Do you get a warning printed before the reload with #24126 applied?

@PVince81
Copy link
Contributor

Setting to critical as this makes the admin page unusable in such environments.

@PVince81
Copy link
Contributor

CC @cmonteroluque FYI

@PVince81
Copy link
Contributor

@LukasReschke any progress on the debugging front ?

@PVince81
Copy link
Contributor

PVince81 commented May 2, 2016

Likely an env issue, moving to 9.0.3 as it would still be good to find out what is going on

@cmonteroluque

@PVince81 PVince81 modified the milestones: 9.0.3-next-maintenance, 9.0.2-current-maintenance May 2, 2016
@PVince81
Copy link
Contributor

PVince81 commented Jun 2, 2016

Ok, so the SSL warning might be causing a reload from what I read

@PVince81
Copy link
Contributor

PVince81 commented Jun 8, 2016

It looks like the global ajax error handler is getting a lot of false positives, and that only because it is assuming that a failed XmlHttpRequest is due to a cross-domain request trying to connect to an IDP in a SSO situation.

Anyway, I think we should backport #24126 which will at least introduce a delay and give the admin a chance to see the current messages, which can already be useful to find out what is wrong in their env.

I'll prepare a backport PR.

@PVince81
Copy link
Contributor

PVince81 commented Jun 8, 2016

PR here #25035

@icewind1991
Copy link
Contributor

@PVince81 is this closed with #25035 ?

@PVince81
Copy link
Contributor

Kind of, yes. At least it gives admin a better chance of seeing warnings and debugging.

However it was reported again here #25058 (comment) maybe with more chance to find out about the core issue.

@fuco809
Copy link

fuco809 commented Jun 21, 2016

same issue here. OC9.0.2.2 + php7 + centos7(latest patches) proxy needed and configured in config.php and additional reverse-proxy apache in front - then issue appears when using http://ocserver but when using https it works. and when accessing directly to owncloud-apache (without reverse proxy) it works too with http (and https).

@fuco809
Copy link

fuco809 commented Jun 21, 2016

reload_site_oc9

after login to OC i get the loop on /owncloud/apps/files/

i tried to figure out the root cause. it seems that js.js makes the reload (see screenshot)
i also tried #25058 tips with apache config and overwritehost/overwritehostport. - no effect.
i also tried OC9.0.3RC1, and there ist the 5sec delay message, but problem still exists.
any hints?

the add content button does not appear (before reload)
no_add_content

@PVince81
Copy link
Contributor

js.js makes the reload because it detected an ajax error and believes that it was a cross-site SSO redirect. Currently there is no good way to detect such situations.

In your specific case you need to find out why the ajax call causes an error in the first place, likely an issue in your env.

@fuco809
Copy link

fuco809 commented Jun 21, 2016

with browser debugging mode i found this message while loading the page, but i am not sure if this is relevant.

The requested URL /core/vendor/jquery/jquery-1.10.2.js was not found on this server.

@n2five6
Copy link

n2five6 commented Jun 22, 2016

We have the same problem OC 9.0.2 behind an LB.
The LB delivers the SSL-Cert (no Cert is configured on server side).

If I try to debug the error with FF, I saw also that there is something with the Content-Security-Policy. Maybe it has also to do with the .ocdata error (that we have also).

owncloud-redirect

[Wed Jun 22 15:08:20.835144 2016] [authz_core:error] [pid 3020] [client ██.██.██.██:12784] AH01630: client denied by server configuration: /srv/www/owncloud-lt/data/.ocdata

@fuco809
Copy link

fuco809 commented Jun 23, 2016

for my case i found the cause and a solution. the propfind http requests are somewhere filtered (IPS,...) and without them the page allways reloads. now after fixing this, the reloads disappear and all working fine.
@n2five6 : as your screenshot shows you have at least the same problem with propfind requests.

@PVince81
Copy link
Contributor

Now thinking of it, I might have a solution to prevent the reload in case the setup checks are triggering the "global ajax error" condition. I'll add an exclusion flag for the setup checks.

@PVince81
Copy link
Contributor

Fix here: #25239

Can you guys who have the reload issue try it out ?

@fuco809
Copy link

fuco809 commented Jun 23, 2016

applied the changes in core/js/setupchecks.js on a 9.0.3rc1 ... now the page reload (loop) dissappears, but OC not fully loading/not functionally because propfind http request times out (on my testserver).

@PVince81
Copy link
Contributor

@fuco809 thanks. Yeah, there is indeed a setup issue on your side.

Goal of my PR was mostly to avoid the annoying reload while trying to figure out how to fix the env.

@n2five6
Copy link

n2five6 commented Jun 24, 2016

@fuco809: Thanks! I looks like there will be something blocked on our LB. We opened now a support Ticket by KEMP.

@PVince81: Is there a List which HTTP-Methods are used from ownCloud?

@PVince81
Copy link
Contributor

I can't seem to find an explicitly list. Mostly the verbs commonly used by REST and Webdav APIs:

  • GET
  • POST
  • PUT
  • MOVE
  • DELETE
  • PROPPATCH
  • PROPFIND
  • REPORT
    (from the top of my head)

@ghost
Copy link

ghost commented Jun 26, 2016

There are also others like MKCOL. Personally i'm using those:

https://forum.owncloud.org/viewtopic.php?f=23&t=10519&p=26878#p26878

@NestorTejero
Copy link

On OwnCloud 9.1.0 we are getting the same error. Our code already contains the fix linked above, but it looks like it is just hiding the error, not going to the root cause.

We get this:

admin:1 Mixed Content: The page at 'https://owncloud.XXX.com/settings/admin' was loaded over HTTPS, but requested an insecure XMLHttpRequest endpoint 'http://owncloud.XXX.com/apps/files/'. This request has been blocked; the content must be served over HTTPS.

Just for reference, we have an SSL certificate running in the server, so that the main page for OC is https://owncloud.XXX.com.

@PVince81
Copy link
Contributor

Yes, the above code only removes the useless reloading to make it possible for you to examine the network console and other possible environment errors.

@lock
Copy link

lock bot commented Aug 3, 2019

This thread has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.

@lock lock bot locked as resolved and limited conversation to collaborators Aug 3, 2019
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.
Projects
None yet
Development

Successfully merging a pull request may close this issue.

7 participants