Page Cannot Be Loaded Error

  • Bug
  • Thread starter Vanadium 50
  • Start date
In summary, the conversation discusses an error message that appears when trying to load pages on the Physics Forums website. The error seems to be server-side and not related to the user's internet connection. Other users have also reported similar issues and it has been suggested that it may be due to an issue with the XenForo JavaScript. The problem seems to be cumulative and affects the usability of the site. Some users have experienced slow internet speeds, but the consensus is that the issue lies with the website rather than the internet connection.
  • #36
Doing it now.
 
Physics news on Phys.org
  • #37
Vanadium 50 said:
Doing it now.
I reached out to a Xenforo dev and he said he's seen this happen for those using an ad blocker and visiting a page with "ad" somewhere in the title or page URL. Could this be the case here?
 
  • #38
I am using an ad-blocker but this goes away after a minute. But do a search for "ad" - you see it in lots of places: Academic Advising.
 
  • #39
Worse - any URL ending in unread will have the AD.
 
  • #40
Vanadium 50 said:
Worse - any URL ending in unread will have the AD.
Even worse "thread" is in every URL.
 
  • #41
Greg Bernhardt said:
I reached out to a Xenforo dev and he said he's seen this happen for those using an ad blocker and visiting a page with "ad" somewhere in the title or page URL. Could this be the case here?
I don't think so. Ad blockers generally do not cause a problem in site loading; in fact, often blocking ads makes a site load much faster. In addition, when we log into PF, no one can see any ads, so ad blockers will not really have an effect here.
 
  • #42
Wrichik Basu said:
I don't think so. Ad blockers generally do not cause a problem in site loading; in fact, often blocking ads makes a site load much faster. In addition, when we log into PF, no one can see any ads, so ad blockers will not really have an effect here.
Whether it's the case here, unsure, but the error message literally addresses ad blockers as a possible cause.
 
  • #43
I'm not so sure the error message is the One True Font of Knowledge here. Lots of things can go into error messages. Why not Haiku?

The page you asked for
Sadly is here no longer
Moved permanently


I'm also not sure we can say "Well, just shut off your ad blocker". These days, pretty much all browsers come with ad blockers. I might even make the argument that the real benefit of ad blockers isn't the reduction of legitimate ads, it's a defense against malvertising. I would certainly want more specific advice, like "whitelist physicsforums" than "shut your protection off - it might help. Then again, it might not"

Finally, we have presented evidence that the problem is server side. The best evidence is that a killing a stuck job made the problem go away temporarily. The next best is that multiple people report it. The next best is that it is intermittent. I don't see any evidence against the theory that the server is overwhelmed with the zillion scripts Xenforo wants to run.
 
  • #44
Happened a minute ago. Do you have enough data points? Can I stop recording?
 
  • #45
Vanadium 50 said:
Happened a minute ago. Do you have enough data points? Can I stop recording?
To be honest, I'm at a loss as to how to diagnose this further. There are no signs the server is taxed. We're not hearing from a large number of people. I reached out to a server admin and he didn't have any ideas. What do you suggest?
 
  • #46
Greg Bernhardt said:
To be honest, I'm at a loss as to how to diagnose this further. There are no signs the server is taxed. We're not hearing from a large number of people. I reached out to a server admin and he didn't have any ideas. What do you suggest?
It would be good if you could recreate the error. Connect via a low-speed public wifi (or effectively simulate such a connection).

For me, it seems to happen mostly on pop-up functionality (trying to edit or reply to a post, or simply like a post, or invoke the search function). Someone would need to look at what the software does precisely in those cases and figure out what could go wrong.

The big clue, I think, is that I almost always get a couple of functions on a new window. Then, it seems to get tangled up (it must be something to do with low bandwidth and/or what happens on public wifi). It seems to me that it loses the place and then after 10-15 seconds times out.
 
  • Informative
Likes Greg Bernhardt
  • #47
Presumably this message is sent in response to severl conditions. Can you make one page for half of the conditions and another for the other half?
 
  • #48
Vanadium 50 said:
Presumably this message is sent in response to severl conditions. Can you make one page for half of the conditions and another for the other half?
Sorry I don't know what this means
 
  • #49
This page gets sent when something happens. Probably a list of things. Can you find that list and split it up? I am thinking along the lines of a binary search through the possibilities.
 
  • #50
Vanadium 50 said:
This page gets sent when something happens. Probably a list of things. Can you find that list and split it up? I am thinking along the lines of a binary search through the possibilities.
I got a little more info from an XF dev. He said this error page is cached by the browser as a requirement to meet PWA standards. Still thinking of how to diagnose. Emulating conditions is not easy especially if it is a weird server hiccup.
 
  • #51
Frustrating! I very occassionally get the same un-informative msg. Here, it usually, but not always, is an Internet problem. The worst part is that msg. seems to be a "last resort, catchall, generic" response.

From both personal experience and much online research time, here are the possibilities/conclusions I've come to, in no particular order:

Likely Causes:
1) A DNS problem
2) Page is truly not available on server due to a server problem
3) A dropped data packet from site to you
4) The communication stack on your computer got confused/corrupted
5) The browser cache is corrupted
6) Insufficient memory when browser asks for more storage
. . (too many programs or tabs open, OS bug?)
7) The browser has a bug
8) A few sites suggested using a VPN1) and 2) are usually corrected within minutes to days.
3) retry
4) reboot
5) and 6) clear cache & restart browser
7) and 8) ??

In my experience, an intermittent bug on the server side seems to be rare. Their usual error message says Page Not Found, not that you couldn't receive or display it.

Cheers,
Tom
 
  • #52
Vanadium 50 said:
Finally, we have presented evidence that the problem is server side. The best evidence is that a killing a stuck job made the problem go away temporarily. The next best is that multiple people report it. The next best is that it is intermittent. I don't see any evidence against the theory that the server is overwhelmed with the zillion scripts Xenforo wants to run.
The servers are certainly involved but it's not exclusively a server issue, or it would affect everyone equally (at the same time). Some users encounter that problem regularly. I haven't seen it even once. This thread is heavily biased towards users who see it, obviously.
 
  • #53
Tom.G said:
Their usual error message says Page Not Found, not that you couldn't receive or display it.
You need to appreciate that what you are seeing is an error message generated by the XenForo JavaScript running on the page when it fails to get a response to what is generally called an XMLHttpRequest. The routing of any requests to physicsforums.com are handled by the Cloudflare CDN which then forwards them to Greg's origin server. Cloudflare has some security (anti-DDoS) software which may be getting in the way, also Cloudflare has recently introduced a feature called 'Rocket Loader' which rewrites every HTML page that it delivers to use a proprietory asynchronous loader.

There are many things that all need to work, and the process is complicated further by browser caching (some of these scripts are delivered with a long cache time which means that unless the user manually clears the browser cache any changes Greg makes at his end, even if he pushes a cache invalidation to Cloudflare, are not going to be picked up locally).

It seems to me that the problems have started happening recently, and recently 2 things have happened:
  • Cloudflare have implemented Rocket Loader
  • PF has updated to a new version of XenForo
Tracking down whether the problem is due to one of these changes, a combination of the two changes, or some other interaction with another part of the system, is not trivial.
 
  • Like
  • Informative
Likes Astronuc, Tom.G, Greg Bernhardt and 1 other person
  • #54
pbuk said:
  • Cloudflare have implemented Rocket Loader
  • PF has updated to a new version of XenForo
Both of these ran across my mind. We do have a minor upgrade to do for XF which I may try to get to this weekend. For the Rocket Loader, we can try turning it off for a week and see if that helps, but generally, it has helped page loading.
 
  • Like
Likes dextercioby and pbuk
  • #55
A fact to add to the list - when it was doing this (well, the pause before the message), I opened a new window and a new private window. The two ""regular" windows displayed the error page. The private window displayed nothing.
 
  • Like
Likes Greg Bernhardt
  • #56
One more data point. I went to the forum home page and it was taking a while to load - often a precursor to this message. I fired up Edge (I usually use Firefox) and immediately got the visitor page. I clicked the Log On button and...nothing. I got the message in the Firefox window, reloaded, and got the page. AND Edge was now displaying the login window.
 
  • #58
Greg Bernhardt said:
Anyone still getting the error?
I'll be on my phone for a few weeks. The problem was only on the laptop.
 
  • Like
Likes Greg Bernhardt
  • #59
Greg Bernhardt said:
Anyone still getting the error?
A few minutes ago.
 

Similar threads

Replies
5
Views
1K
Replies
5
Views
1K
Replies
0
Views
96K
Replies
3
Views
1K
Replies
6
Views
1K
Replies
35
Views
6K
Replies
13
Views
3K
Replies
15
Views
5K
Back
Top