Bug 1940993

Summary: IPv6 problems after upgrading to Fedora 34
Product: [Fedora] Fedora Reporter: Dima Ryazanov <dima>
Component: systemdAssignee: systemd-maint
Status: CLOSED DUPLICATE QA Contact: Fedora Extras Quality Assurance <extras-qa>
Severity: high Docs Contact:
Priority: unspecified    
Version: 34CC: bugzilla.redhat.com, fedoraproject, filbranden, flepied, kasong, lnykryn, msekleta, ssahani, s, systemd-maint, yuwatana, zbyszek
Target Milestone: ---   
Target Release: ---   
Hardware: x86_64   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: If docs needed, set a value
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2021-03-23 17:19:40 UTC Type: Bug
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Attachments:
Description Flags
`strace curl` of the failing request none

Description Dima Ryazanov 2021-03-19 17:49:16 UTC
Description of problem:

After upgrading to Fedora 34, about half of the time, I cannot connect to IPv6-enabled websites. E.g.:

$ curl -v 'https://mail.google.com/mail/u/0/'
*   Trying 2607:f8b0:4005:809::2005:443...
* Immediate connect fail for 2607:f8b0:4005:809::2005: Network is unreachable
* Closing connection 0
curl: (7) Couldn't connect to server

The DNS name resolves to both IPv4 and IPv6:

$ host mail.google.com
mail.google.com is an alias for googlemail.l.google.com.
googlemail.l.google.com has address 172.217.5.101
googlemail.l.google.com has IPv6 address 2607:f8b0:4005:801::2005

I don't have IPv6 routes:

$ route -n
Kernel IP routing table
Destination     Gateway         Genmask         Flags Metric Ref    Use Iface
0.0.0.0         192.168.87.1    0.0.0.0         UG    600    0        0 wlp1s0
192.168.87.0    0.0.0.0         255.255.255.0   U     600    0        0 wlp1s0

I'm not sure if IPv6 has ever actually worked in my network - but I've never run into problems in Fedora 33, so I'm assuming it did not try to use IPv6.

(Not sure where the bug is coming from; just guessing it might be systemd-resolved, so assigning to it.)

Version-Release number of selected component (if applicable):


How reproducible:

About half of the time; looks like it picks IPv4 or IPv6 address at random.

Steps to Reproduce:
1. Connect to mail.google.com, use "dnf update", etc.
2.
3.

Actual results:
Network is unreachable

Expected results:
Use IPv4

Additional info:

Comment 1 Dima Ryazanov 2021-03-19 17:51:57 UTC
*** Bug 1940994 has been marked as a duplicate of this bug. ***

Comment 2 ojab 2021-03-20 21:54:27 UTC
It started happening here rather recently (I'm on F34 for a month already) and AFAIR only related updates before it became broken are glibc/systemd.

I'm updating the system via `dnf distro-sync --refresh` from the console and in the last week started getting 
```
Errors during downloading metadata for repository 'google-cloud-sdk':
  - Curl error (7): Couldn't connect to server for https://packages.cloud.google.com/yum/doc/yum-key.gpg []
  - Curl error (7): Couldn't connect to server for https://packages.cloud.google.com/yum/repos/cloud-sdk-el7-x86_64/repodata/repomd.xml.asc []
Error: Failed to retrieve GPG key for repo 'google-cloud-sdk': Curl error (7): Couldn't connect to server for https://packages.cloud.google.com/yum/doc/yum-key.gpg []
```
and similar errors for other repos, but taken a look only today why it's happening since `mpv --ytdl` was failing and youtube was working for sure in the browser.

Comment 3 ojab 2021-03-20 21:55:39 UTC
Created attachment 1764953 [details]
`strace curl` of the failing request

`strace -ff -o /tmp/st/curl -s 100000 curl youtube.com`

Comment 4 Dima Ryazanov 2021-03-22 19:16:14 UTC
Just ran into the same problem while connected to a different network, so more evidence this isn't network-specific.

Also...

$ curl -v 'https://mirrors.fedoraproject.org/metalink?repo=fedora-modular-34&arch=x86_64'
*   Trying 2620:52:3:1:dead:beef:cafe:fed7:443...
* Immediate connect fail for 2620:52:3:1:dead:beef:cafe:fed7: Network is unreachable
*   Trying 2620:52:3:1:dead:beef:cafe:fed6:443...
* Immediate connect fail for 2620:52:3:1:dead:beef:cafe:fed6: Network is unreachable
*   Trying 2604:1580:fe00:0:dead:beef:cafe:fed1:443...
* Immediate connect fail for 2604:1580:fe00:0:dead:beef:cafe:fed1: Network is unreachable
*   Trying 2605:bc80:3010:600:dead:beef:cafe:feda:443...
* Immediate connect fail for 2605:bc80:3010:600:dead:beef:cafe:feda: Network is unreachable
*   Trying 2610:28:3090:3001:dead:beef:cafe:fed3:443...
* Immediate connect fail for 2610:28:3090:3001:dead:beef:cafe:fed3: Network is unreachable
*   Trying 2605:bc80:3010:600:dead:beef:cafe:fed9:443...
* Immediate connect fail for 2605:bc80:3010:600:dead:beef:cafe:fed9: Network is unreachable
* Closing connection 0
curl: (7) Couldn't connect to server


"dead:beef:cafe"? Is this for real?

Comment 5 ojab 2021-03-22 19:25:16 UTC
Some more bugzilla searching shows https://bugzilla.redhat.com/show_bug.cgi?id=1940715, I suppose this one can be closed.

Comment 6 Zbigniew Jędrzejewski-Szmek 2021-03-23 17:19:40 UTC
Yeah, this should be the same issue. Please test with the latest update.

*** This bug has been marked as a duplicate of bug 1940715 ***