"Unable to connect" while Internet connection is operational
Categories
(Core :: Networking, defect)
Tracking
()
People
(Reporter: u20230201, Unassigned)
Details
User Agent: Mozilla/5.0 (X11; Linux x86_64; rv:91.0) Gecko/20100101 Firefox/91.0
Steps to reproduce:
Using my laptop with a wired (DHCP) connection, I suspended it over night (as many nights before). Next morning tabs already open could not connect.
I had checked the IP on eth0, and it was OK. A ping to google also succeeded.
The nameservers also looked OK and they worked.
Also the w3m browser could connect, but Firefox could not.
Eventually restarting Firefox resolved the issue.
One issue I know from the past is when you keep firefox open, but move to a different network where DHCP may assign a different hostname/IP. The Firefox was also unable to connect until it was restarted.
Actual results:
Firefox error message:
Unable to connect
Firefox can’t establish a connection to the server at www.google.com.
The site could be temporarily unavailable or too busy. Try again in a few moments.
If you are unable to load any pages, check your computer’s network connection.
If your computer or network is protected by a firewall or proxy, make sure that Firefox is permitted to access the Web.
Expected results:
There should be a better diagnostic message at least.
|
||
Comment 1•4 years ago
|
||
The Bugbug bot thinks this bug should belong to the 'Core::Networking' component, and is moving the bug to that component. Please revert this change in case you think the bot is wrong.
|
Reporter | |
Comment 2•4 years ago
|
||
Maybe I should add that the error message was displayed immediately (and not after some significant delay).
|
||
Comment 3•4 years ago
|
||
Could you try to record a http log when this problem happens? Thanks.
|
|
Reporter | |
Comment 4•4 years ago
|
||
(In reply to Kershaw Chang [:kershaw] from comment #3)
Could you try to record a http log when this problem happens? Thanks.
As the method will require a browser restart (to get the environment variables set), it wouldn't really help as a browser erstart fixed the problem.
Also I had opened the network console when the problem was there, and I didn't see anything other than the request being displayed.
|
|
||
Comment 5•4 years ago
|
||
(In reply to Ulrich Windl from comment #4)
(In reply to Kershaw Chang [:kershaw] from comment #3)
Could you try to record a http log when this problem happens? Thanks.
As the method will require a browser restart (to get the environment variables set), it wouldn't really help as a browser erstart fixed the problem.
Also I had opened the network console when the problem was there, and I didn't see anything other than the request being displayed.
You could use about:networking#logging to start logging without restarting Firefox.
Thanks.
|
|
||
Comment 6•4 years ago
|
||
Close because we don't have enough information.
Feel free to reopen if you have a http log. Thanks.
|
|
Reporter | |
Comment 7•4 years ago
|
||
Using MozillaFirefox-91.8.0-150200.152.26.1.x86_64 from OpenSUSE Leap 15.3 the problem hit me agian:
I had multiple tabs open when I tried to open another (assumed safe) website on another tab. That page did not open, but I noticed some content flickering and the CPU fan started to spin up. So after some time I closed that tab. I think I could continue to work (but that was yesterday).
So I closed the laptop lid and opened it again this morning when I wanted to lookup something.
The page wouldn't load, and the CPU fan was spinning up again. In fact none of the tabs would load.
Running an strace on the most busy process showed that no system calls were made while 100% CPU was consumed.
At that point I realized that any URL to help debugging the issue was rather useless 8-(
So I tried a bit and started network logging for a while. A lot of lines were logged withing short time, so I stopped it again soon.
I decided to stop Firefox and restart it, restoring the history. As I'm typing this on the very same machine using firefox, it seems to work again ;-)
When closing Firefox I had noticed that a process was still consuming 100% CPU for some time before it was closed (and the CPU fan came down again):
Tasks: 361 total, 3 running, 356 sleeping, 2 stopped, 0 zombie
%Cpu(s): 14.1 us, 0.7 sy, 0.0 ni, 84.9 id, 0.0 wa, 0.0 hi, 0.3 si, 0.0 st
MiB Mem : 15219.04+total, 8343.098 free, 3221.988 used, 3653.961 buff/cache
MiB Swap: 6141.996 total, 6141.996 free, 0.000 used. 11401.20+avail Mem
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
4093 wiuxxxxx 20 0 3164432 270956 143968 R 100.7 1.739 89:43.20 Web Content
3677 wiuxxxxx 20 0 5785380 983.2m 488832 S 15.95 6.460 25:43.47 firefox
2912 wiuxxxxx 20 0 1594848 107520 66496 S 2.326 0.690 5:05.40 X
...
(The machine has 4 cores, two threads each, so 14% CPU usage is more than one thread running at 100%)
I also tried to ping one of the names that would not respond, and the ping just worked fine.
ping www.stackexchange.com
PING stackexchange.com (151.101.65.69) 56(84) bytes of data.
64 bytes from 151.101.65.69 (151.101.65.69): icmp_seq=1 ttl=58 time=8.61 ms
64 bytes from 151.101.65.69 (151.101.65.69): icmp_seq=2 ttl=58 time=8.61 ms
64 bytes from 151.101.65.69 (151.101.65.69): icmp_seq=3 ttl=58 time=8.57 ms
^C
--- stackexchange.com ping statistics ---
3 packets transmitted, 3 received, 0% packet loss, time 2003ms
|
|
Reporter | |
Updated•4 years ago
|
|
|
Reporter | |
Comment 8•4 years ago
|
||
Browsing the network log file I noticed that it contains confidential information like authentication cookies in headers and URL parameters (not to talk about others), so I do not want to publish it as attachment here. Sorry. Any solutions for that?
|
|
||
Comment 9•4 years ago
|
||
(In reply to Ulrich Windl from comment #8)
Browsing the network log file I noticed that it contains confidential information like authentication cookies in headers and URL parameters (not to talk about others), so I do not want to publish it as attachment here. Sorry. Any solutions for that?
Could you send the log file to necko at mozilla dot com?
Thanks.
|
|
||
Updated•4 years ago
|
|
|
Reporter | |
Comment 10•4 years ago
|
||
(In reply to Kershaw Chang [:kershaw PTO until July 25] from comment #9)
Could you send the log file to
necko at mozilla dot com?
I want to apologize: Somehow I missed your message, and when looking for it, it seems I have removed the file in the meantime.
|
||
Comment 11•4 years ago
|
||
Without http logs, not much we can do here so I am going to close.
Please re-open again if you are able to reproduce with http logs.
Description
•