[Netalyzr] IPv6 (sometimes) broken?
nweaver at ICSI.Berkeley.EDU
Sat Jul 18 06:53:13 PDT 2015
> On Jul 18, 2015, at 5:16 AM, Bjoern A. Zeeb <bzeeb-lists at lists.zabbadoz.net> wrote:
> we have users at IETF trying netalyzer and getting bogus results for IPv6, e.g.,
> The resolver lines, but also “IPv6 connectivity (?): No IPv6 support” seem wrong. IPv6 support works, and the resolvers do answer as we just tested.
> Are there some firewall or connectivity issues on the netalyzer server ends? We had that before.
> I did re-test myself and didn’t see errors but I couldn’t spot from where individual tests were run and it could just be a 1 in n issue?
This test is test-14 in the transcript (checkV6):
023.896 test-14| Starting checkV6
024.858 test-14| Checking TCP for tcpServer: ipv6-node.u6880.n1.netalyzr.icsi.berkeley.edu/2607:f740:b:0:0:0:0:f93
024.861 test-14| connecting to /2607:f740:b:0:0:0:0:f93:1947: unavailable
024.861 test-14| Local address is null
024.861 test-14| Remote adddress is null
024.861 test-14| Data received:
024.861 test-14| Now checking for V4 server: ipv4-node.u6880.n1.netalyzr.icsi.berkeley.edu/18.104.22.168
024.959 test-14| Connected to /22.214.171.124:1947 in 98 ms
024.959 test-14| reading 1200 bytes from socket...
025.056 test-14| Local address is 126.96.36.199
025.056 test-14| Remote adddress is 188.8.131.52
025.056 test-14| Data received: %3A%3Affff%3A184.108.40.206%2155274
We have had issues with the server in the past, but its working right now:
nweaver% telnet 2607:f740:b:0:0:0:0:f93 1947
Connected to ipv6-node.netalyzr.icsi.berkeley.edu.
Escape character is '^]'.
2601:643:1:b27:ed0d:a9b9:af8e:844a!58228Connection closed by foreign host.
nweaver% telnet ipv6-node.u6880.n1.netalyzr.icsi.berkeley.edu 1947
Connected to ipv6-node.u6880.n1.netalyzr.icsi.berkeley.edu.
Escape character is '^]'.
2601:643:1:b27:ed0d:a9b9:af8e:844a!58236Connection closed by foreign host.
But I wonder if there is some issue with Java or the host on this system: EG, the IPv6 allocated DNS servers are also failing to be contacted:
102.959 test-37| Collecting data on resolver 2001:67c:370:229::7
102.959 test-37| Performing DNS fetch check of query server.u6880.n1.netalyzr.icsi.berkeley.edu
102.959 test-37| To server 2001:67c:370:229::7
102.959 test-37| With EDNS0 MTU of 0 and want_dnssec of false
102.959 test-37| Query to server is 60 bytes
102.959 test-37| Testing query as DNS to make sure it parses
102.959 test-37| DNS: id: 6169 qr: false opcode: 0
102.959 test-37| DNS: aa: false tc: false rd: true ra: false
102.959 test-37| DNS: rcode: 0
102.959 test-37| DNS: Questions:
102.959 test-37| DNS: server.u6880.n1.netalyzr.icsi.berkeley.edu QTYPE: 1 QCLASS: 1
102.959 test-37| DNS: Answers:
102.959 test-37| DNS: Authority:
102.959 test-37| DNS: Additional:
102.960 test-37| Error in querying, caught: java.io.IOException: No route to host
I wonder if this is related to the tunnel interfaces: there are two V6-only tunnel interfaces, awdl0 and utun0 on the system this test was run from (Local Network Interfaces). The first is an Apple P2P one, but the second might be the source of the issue?
What happens if you try a direct connection to known good Ipv6 only sites like ipv6.google.com?
Nicholas Weaver it is a tale, told by an idiot,
nweaver at icsi.berkeley.edu full of sound and fury,
510-666-2903 .signifying nothing
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 842 bytes
Desc: Message signed with OpenPGP using GPGMail
Url : http://mailman.ICSI.Berkeley.EDU/pipermail/netalyzr/attachments/20150718/212a35fd/attachment.bin
More information about the Netalyzr