AS64445 - AS-NETJOIN NetJoin srlPI/PARFC 4890 ✓RPKI ✓ 5/5ASPA -

AS overview

Test target: 2a12:e6c0::2 · www.netjoin.it · address space: mixed · prefixes announced: 5

prefix(es): 2a12:e6c0:1::/48, 2a04:68c0::/32, 2a12:e6c0:1337::/48, 2a12:e6c0::/48, 2a12:e6c0:c50::/48

MIX-IT / NAMEX / MINAP / TOP-IX / VSIX / EQUINIX-MILAN / BGPX-ROME / NAMEX-BARI / PCIX / DECIX-PALERMO peering

Random-IP probe (yarrp) into the AS' prefixes

We probed 30 arbitrary IPv6 address(es) inside AS64445's announced prefix(es). No router inside the prefix responded. The deepest visible hop was 2400:54a0:1010::210 at hop 18 - that router answers ICMPv6 Echo. This is not in AS64445's announced prefix; operationally it's the AS-edge / peering interface (often on an IXP peering address or an upstream's /127 link). The traceroute boundary is the AS edge: ICMPv6 Time-Exceeded responses from anything inside this AS are filtered.

ICMPv6 Type 2 (Packet Too Big) acceptance - active test

Vantage points disagree about Type 2 acceptance - transit ASes on one path may be filtering Type 2 even though the destination's stack accepts it on another path:

• NL host (go6lab): Type 2 honored. Forged PTB accepted; next response shrunk. (ICMPv6 Echo + forged PTB)
• ITA host (Karsolink): Type 2 NOT honored. Forged PTB had no effect (filtered en route or stack ignored it). (TLS handshake + forged PTB)
• SLO host (6connect): Type 2 honored. Forged PTB accepted; next response shrunk. (ICMPv6 Echo + forged PTB)

Why they disagree - different probe methods: These vantages picked different probe methods, so their verdicts are not directly comparable. The active test tries methods in order (icmp6-echo → dns-tcp → tls → http) and stops at the first that produces a definitive verdict; if the chosen method differs, the underlying evidence differs too. Most often this is because the path PMTU on one vantage is below 1500 B, so the natural Echo Reply is already fragmented and the icmp6-echo method falls through to a TCP-based one. This is largely a measurement artifact, not a destination-behaviour difference.

sudo tcpdump -i any -n -e 'icmp6 and ip6[40] = 2 and (src host 2607:fae0:a000::42 or src host 2a00:8642:42::75 or src host 2a12:d8c0:105a:9001::a154)'

sudo ip6tables -L INPUT -nv | grep -iE 'icmpv6|packet-too-big'
sudo nft list ruleset 2>/dev/null | grep -A1 'icmpv6'

sudo sysctl net.ipv6.conf.all.accept_redirects net.ipv4.ip_no_pmtu_disc net.ipv6.route.mtu_expires

sudo tcpdump -i any -n -e 'icmp6 and ip6[40] = 2 and (src host 2607:fae0:a000::42 or src host 2a00:8642:42::75 or src host 2a12:d8c0:105a:9001::a154)'

From NL host (go6lab)   openRFC 4890 ✓

filter likely at: (none) · min PMTU on path: 1500

Path (traceroute + mtr + PTR)

tracepath -6

 1?: [LOCALHOST]                        0.038ms pmtu 1500
 1:  2a00:8642:42::3                                       1.941ms 
 1:  2a00:8642:42::3                                       1.493ms 
 2:  no reply
 3:  no reply
 4:  be1.core3.ams1.he.net                                 5.693ms 
 5:  be2.core4.fra1.he.net                                10.965ms 
 6:  no reply
 7:  no reply
 8:  no reply
 9:  netjoin.minap.it                                     18.832ms 
10:  2a00:8c60:c:1::1                                     18.173ms 
11:  2a12:e6c0::2                                         18.675ms reached
     Resume: pmtu 1500 hops 11 back 11 

From ITA host (Karsolink)   openRFC 4890 ✗

filter likely at: (none) · min PMTU on path: 1500

Path (traceroute + mtr + PTR)

Rate-limited ICMPv6: hop 3 (loss between 5% and 95% across mtr cycles - the router replies but only sometimes).

tracepath -6

 1?: [LOCALHOST]                        0.030ms pmtu 1500
 1:  karsolink-01.net.karsolink.com                        0.556ms 
 1:  karsolink-01.net.karsolink.com                        0.986ms 
 2:  2a12:d8c0:109f:121::a1                                0.783ms 
 3:  2a12:d8c0:101f:6::1                                  10.403ms 
 4:  netjoin.minap.it                                     10.433ms 
 5:  2a12:e6c0::1                                         10.552ms 
 6:  2a12:e6c0::2                                         10.886ms reached
     Resume: pmtu 1500 hops 6 back 6 

From SLO host (6connect)   openRFC 4890 ✓

filter likely at: (none) · min PMTU on path: 1500

Path (traceroute + mtr + PTR)

Rate-limited ICMPv6: hop 6 (loss between 5% and 95% across mtr cycles - the router replies but only sometimes).

tracepath -6

 1?: [LOCALHOST]                        0.037ms pmtu 1500
 1:  fw1-lju.6connect.com                                  0.555ms 
 1:  fw1-lju.6connect.com                                  0.514ms 
 2:  no reply
 3:  2a03:a100:0:201:1::1                                 76.748ms 
 4:  e0-1.core1.lju1.he.net                                2.013ms 
 5:  no reply
 6:  port-channel1.core1.zag2.he.net                       4.566ms 
 7:  no reply
 8:  netjoin.minap.it                                     12.265ms 
 9:  2a06:1287:330d:6445::2                               12.636ms 
10:  2a06:1287:330d:6445::2                               12.126ms asymm  9 
11:  2a12:e6c0::2                                         13.682ms reached
     Resume: pmtu 1500 hops 11 back 10