2

I just started using a pi2 v1.1. It's old, I know. I'd buy and use a 4, but can't get my hands on one.

I loaded the pi2 with a fresh raspbian (bullseye) lite using the imager. I updated everything with an apt get.

When I ping, I get strange behavior, depending on the site I ping. Ping doesn't work for big sites like yahoo, Google, or Microsoft. For example,

 $ ping www.yahoo.com
 PING www.yahoo.com(media-router-fp73.prod.media.vip.bf1.yahoo.com (2001:4998:124:1507::f000)) 56 data bytes
 ^C
 --- www.yahoo.com ping statistics ---
 3 packets transmitted, 0 received, 100% packet loss, time 57ms

Google and Microsoft have similar behavior where the ping doesn't go to an IP address but rather another domain. However, ping does work as expected for other sites, e.g.

 $ ping www.longform.org
 PING longform.org (165.227.71.159) 56(84) bytes of data.
 64 bytes from 165.227.71.159 (165.227.71.159): icmp_seq=1 ttl=43 time=39.8 ms
 64 bytes from 165.227.71.159 (165.227.71.159): icmp_seq=2 ttl=43 time=39.3 ms
 64 bytes from 165.227.71.159 (165.227.71.159): icmp_seq=3 ttl=43 time=39.0 ms
 ^C
 --- longform.org ping statistics ---
 3 packets transmitted, 3 received, 0% packet loss, time 5ms
 rtt min/avg/max/mdev = 38.996/39.364/39.821/0.342 ms

When I ping using my laptop, I get the expected behavior for yahoo.com, google.com, etc. For example:

$ ping www.yahoo.com
PING new-fp-shed.wg1.b.yahoo.com (74.6.231.21): 56 data bytes
64 bytes from 74.6.231.21: icmp_seq=0 ttl=48 time=57.075 ms

The output of resolv.conf is

domain home
nameserver 192.168.1.1
nameserver 2607:f428:ffff:ffff::2
nameserver 2607:f428:ffff:ffff::1

I've tried the following on the pi:

--changed the /etc/dhcpcd.conf to explicitly include Cloudfare DNS. The resolv.confis now

domain home
nameserver 8.8.4.4
nameserver 8.8.8.8
nameserver 2607:f428:ffff:ffff::2
nameserver 2607:f428:ffff:ffff::1

No dice.

--installed the legacy buster lite. No dice. Also changed resolv.conf on the buster lite. Still no dice.

Any help would be greatly appreciated!

EDIT 1) dig appears to give the correct information:

$ dig www.yahoo.com

; <<>> DiG 9.11.5-P4-5.1+deb10u8-Raspbian <<>> www.yahoo.com ;; global options: +cmd ;; Got answer: ;; ->>HEADER<<- opcode: QUERY, status: NOERROR, id: 58321 ;; flags: qr rd ra; QUERY: 1, ANSWER: 5, AUTHORITY: 0, ADDITIONAL: 1

;; OPT PSEUDOSECTION: ; EDNS: version: 0, flags:; udp: 512 ;; QUESTION SECTION: ;www.yahoo.com. IN A

;; ANSWER SECTION: www.yahoo.com. 31 IN CNAME new-fp-shed.wg1.b.yahoo.com. new-fp-shed.wg1.b.yahoo.com. 31 IN A 74.6.231.21 new-fp-shed.wg1.b.yahoo.com. 31 IN A 74.6.143.26 new-fp-shed.wg1.b.yahoo.com. 31 IN A 74.6.231.20 new-fp-shed.wg1.b.yahoo.com. 31 IN A 74.6.143.25

;; Query time: 24 msec ;; SERVER: 8.8.4.4#53(8.8.4.4) ;; WHEN: Tue Dec 27 15:11:24 GMT 2022 ;; MSG SIZE rcvd: 138

And, just to confirm, ping still doesn't work properly:

$ ping www.yahoo.com
PING www.yahoo.com(media-router-fp74.prod.media.vip.ne1.yahoo.com (2001:4998:44:3507::8001)) 56 data bytes
^C
--- www.yahoo.com ping statistics ---
3 packets transmitted, 0 received, 100% packet loss, time 118ms

$ ping 74.6.231.21 PING 74.6.231.21 (74.6.231.21) 56(84) bytes of data. 64 bytes from 74.6.231.21: icmp_seq=1 ttl=48 time=60.2 ms

EDIT 2) Going to the full raspbian (bullseye) "Raspberry Pi OS with desktop" gives no change in behavior

EDIT 3) Using systemd-networkd installed by following https://raspberrypi.stackexchange.com/a/108593/151115 led to a very, very slow, but ultimately successful ping for www.longform.org and the same behavior as before for www.yahoo.com

EDIT 4) altering resolv.conf to only Cloudflare or Google still led to weird non-IP addresses being used for yahoo.com, etc.

EDIT 5) using the 2015-01-31 raspian OS resolves the issue. I don't know what it is about this distro that makes the problem go away. I'd really rather not use such old an old OS (for security reasons). Any ideas would be much appreciated!

EDIT 6) as requested by @Milliways, the result of ip a is

1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
   valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
link/ether b8:27:eb:f2:57:12 brd ff:ff:ff:ff:ff:ff
inet 192.168.1.243/24 brd 192.168.1.255 scope global dynamic noprefixroute eth0
   valid_lft 258451sec preferred_lft 226051sec
3: tun0: <POINTOPOINT,MULTICAST,NOARP,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UNKNOWN group default qlen 500
link/none 
inet 10.130.173.1/24 scope global tun0
   valid_lft forever preferred_lft forever
4: wg0: <POINTOPOINT,NOARP,UP,LOWER_UP> mtu 1420 qdisc noqueue state UNKNOWN group default qlen 1000
link/none 
inet 10.255.128.1/24 scope global wg0
   valid_lft forever preferred_lft forever

and the result of ip r is

default via 192.168.1.1 dev eth0 proto dhcp src 192.168.1.243 metric 202 
10.130.173.0/24 dev tun0 proto kernel scope link src 10.130.173.1 
10.255.128.0/24 dev wg0 proto kernel scope link src 10.255.128.1 
192.168.1.0/24 dev eth0 proto dhcp scope link src 192.168.1.243 metric 202 

The above is with my fully updated bullseye setup with IPv6 disabled as described in my comment below @Milliway's answer. Current output of resolv.conf is

domain home
nameserver 192.168.1.1
  • "non-IP addresses being used for yahoo.com" 2001:4998:124:1507::f000 is an IP address ... it's an IPv6 address, whereas your laptop seems to only use IPv4 – Jaromanda X Dec 27 '22 at 20:46
  • You STILL haven't explained HOW YOU setup networking. This is obviously NOT any standard setup; you are connected to 2 networks, have some kind of tunnel and are using Ethernet as default. There is no evidence anything is using IPv6 – Milliways Dec 29 '22 at 07:40

1 Answers1

1

You claim to have tried several things (but not specified details of any) and have posted no diagnostics and omitted details of your network provider and how you have configured Pi networking.

Almost certainly the issue is that the Pi is attempting to use IPV6.

On very old OS IPV6 was not enabled by default and the ancient OS you listed uses Debian networking.

You can force ping to to use IPv4 with -4 and can configure dhcpcd to use IPv4 only.

Milliways
  • 59,890
  • 31
  • 101
  • 209
  • Hi @Milliways. Thank you very much for your response. You were right: forcing IPv4 only following @AlexOnLinux's comment in https://raspberrypi.stackexchange.com/questions/23884/force-ipv4-address resolves the issue for me in bullseye. Can you please help me understand how to get IPv6 to work properly? What diagnostics (and what commands to use to get them) would help? The network provider is Spectrum. In terms of configuring the network, the results I showed from the ping's were right out of the box using bullseye-lite downloaded RPi HQ and fresh flashed onto an SD using their installer – astronautgravity Dec 29 '22 at 04:20
  • @WAH the FIRST thing you need to do if you want help is to list (in your Question) HOW YOU setup networking. List the output of ip a && ip r. – Milliways Dec 29 '22 at 05:14
  • 1
    I don't actually use IPv6 (except on my home network) because none of the commercial ISP offer an IPv6 service in Australia so my knowledge is theoretical. – Milliways Dec 29 '22 at 05:17
  • Hi @Milliways, thank you very much again for your time and help. I've edited my question with the output of ip a and ip r – astronautgravity Dec 29 '22 at 06:55