Dreamhost hijacking my prefix...

Not sure how widespread their "leakage" may be, but Dreamhost just
hijacked one of my prefixes...

====================================================================
Possible Prefix Hijack (Code: 10)

Your prefix: 150.182.192.0/18:
Update time: 2013-01-11 14:14 (UTC)
Detected by #peers: 11
Detected prefix: 150.182.208.0/20
Announced by: AS26347 (DREAMHOST-AS - New Dream Network, LLC)
Upstream AS: AS42861 (PRIME-LINE-AS JSC "Prime-Line")
ASpath: 8331 42861 42861 42861 26347

Anyone have a contact there? ASinfo gives netops@dreamhost.com where I
have submitted a report, but so far no joy...

Jeff

Jeff,

We are not announcing the prefix in question nor do we peer with AS42861.

Robtex would beg to differ... you show peered with AS42861, perhaps
someone (else) is looping their advertisements?

_R_egistered
_O_ther side
_B_GP visible Peer
OB AS174 COGENT /PSI
B AS4323 TWTC Autonomous system for tw telecom .
B AS4826 VOCUS-BACKBONE-AS Vocus Connect International Backbone Vocus
Communications Level 2, Vocus House 189 Miller Street North Sydney NSW 2060
B AS5580 ATRATO-IP / Atrato IP Networks
B AS6461 MFNX MFN - Metromedia Fiber Network
B AS6939 HURRICANE Electric
B AS7575 AARNET-AS-AP Australia's Research and Education Network (AARNet3)
B AS7922 COMCAST-IBONE Comcast Cable Communications, Inc. 1800 Bishops
Gate Blvd Mt Laurel, NJ 08054 US
B AS8359 MTS Dummy description for
B AS10912 INTERNAP-BLK Internap Network Services
B AS10913 INTERNAP-BLK Internap Network Services
B AS12989 HWNG Eweka Internet Services B.V.
B AS36351 SOFTLAYER Technologies Inc.
B AS42861 PRIME-LINE-AS Dummy description for

Sounds like someone in Russia is having some fun with as-path prepending and prefix hijacking.

Just checked all BGP speakers again and I show no peering with AS42861.

That would be my guess. We have had some issues with this in the past with
operators from China and Russia.

Here at/as AS5580 I no longer see it announced as a /20, only your own /18:

#sh ip bgp routes 150.182.192.0 255.255.192.0 longer-prefixes
Number of BGP Routes matching display condition : 4
Searching for matching routes, use ^C to quit...
Status A:AGGREGATE B:BEST b:NOT-INSTALLED-BEST C:CONFED_EBGP D:DAMPED
        E:EBGP H:HISTORY I:IBGP L:LOCAL M:MULTIPATH m:NOT-INSTALLED-MULTIPATH
        S:SUPPRESSED F:FILTERED s:STALE
        Prefix Next Hop MED LocPrf Weight Status
1 150.182.192.0/18 80.94.64.10 400 0 BMI
          AS_PATH: 11164 10490 3450 14209
2 150.182.192.0/18 80.94.64.10 400 0 MI
          AS_PATH: 11164 10490 3450 14209
3 150.182.192.0/18 80.94.64.10 400 0 MI
          AS_PATH: 11164 10490 3450 14209
4 150.182.192.0/18 80.94.64.10 400 0 MI
          AS_PATH: 11164 10490 3450 14209

Hi,
Here's a quick summary of what we saw at BGPMon.net.

At 2013-01-11 14:14:13 we saw announcements (seemingly) originated by
26347, for prefixes normally announced by other ASn's (origin change /
hijack).

This seems to have affected 112 prefixes for 110 ASn's [1], including
Rogers, Tata, Sprint, Ziggo, Verizon, KPN, Vodafone, CloudFlare, XS4ALL,
AT&T, Bell Canada and many more.
Most of these were new more specific(!) announcements.

With regards to next-hop ASN's (peers). It seems this hijack was
propagated via 12 unique (AS26347) peers [1]

A quick look at the prefix that was mentioned by Jeff, 150.182.208.0/20
(more specific of 50.182.192.0/18)
The first announcement for this prefix was seen at 2013-01-11 14:14:28
and withdrawn at 2013-01-11 15:20:57. It was detected by 42 unique peers.

some example paths:
271 6939 26347
5580 26347|
37312 5713 6939 26347
1126 24785 12989 26347

[1] I've posted some details (Unique next-hop ASN's and affected origin
ASN's), check if your AS was affected here:
http://portal.bgpmon.net/data/hijack20130111.txt

Cheers,
Andree

.-- My secret spy satellite informs me that at 2013-01-11 7:23 AM Jeff
Kell wrote:

Jeff:

150.182.208.0/20 is not visible from AS702 in Germany.
150.182.192.0/18 path is 702 701 209 26827 14209

Tony

Thanks for that info Andree. The only valid peer I see on the list would
be HE. We do not peer with any of the others listed.

Kenneth

Hi Kenneth,

.-- My secret spy satellite informs me that at 2013-01-11 8:54 AM
Kenneth McRae wrote:

Thanks for that info Andree. The only valid peer I see on the list
would be HE. We do not peer with any of the others listed.

Could it be these ASns receive your routes via an IX route-server?

Below some examples that show a peering between 26347 and
5580 as well as 12989

5580 26347
http://www.ris.ripe.net/cgi-bin/lg/index.cgi?rrc=RRC031&query=12&arg=5580+26347

12989 26347:
http://www.ris.ripe.net/cgi-bin/lg/index.cgi?rrc=RRC031&query=12&arg=12989+26347

And route views:

route-views>sh ip bgp regex 12989_26347
BGP table version is 427410275, local router ID is 128.223.51.103
Status codes: s suppressed, d damped, h history, * valid, > best, i -
internal,
              r RIB-failure, S Stale
Origin codes: i - IGP, e - EGP, ? - incomplete

   Network Next Hop Metric LocPrf Weight Path
* 64.111.96.0/19 208.74.64.40 0 19214 12989
26347 i
* 66.33.192.0/19 208.74.64.40 0 19214 12989
26347 i
* 67.205.0.0/18 208.74.64.40 0 19214 12989
26347 i
* 69.163.128.0/17 208.74.64.40 0 19214 12989
26347 i
* 75.119.192.0/19 208.74.64.40 0 19214 12989
26347 i
* 173.236.128.0/17 208.74.64.40 0 19214 12989
26347 i
* 205.196.208.0/20 208.74.64.40 0 19214 12989
26347 i
* 208.97.128.0/18 208.74.64.40 0 19214 12989
26347 i
* 208.113.128.0/17 208.74.64.40 0 19214 12989
26347 i
* 208.113.200.0 208.74.64.40 0 19214 12989
26347 i

Cheers,
Andree

Hi all,

Atrato / 5580 here.

We don't have direct peering with AS26347, although we learn the AS26347 prefixes through the 206.223.143.253 (AS 19996) routeserver in LAX.

So in a sense we are peering :slight_smile:

Kind regards,

Job

Yes, now that is possible (just no direct peering). So that takes me back
to my original statement about not announcing the 150.182.208.0/20 prefix
to begin with.

Kenneth

.-- My secret spy satellite informs me that at 2013-01-11 10:44 AM
Kenneth McRae wrote:

Yes, now that is possible (just no direct peering). So that takes me
back to my original statement about not announcing the 150.182.208.0/20
<http://150.182.208.0/20&gt; prefix to begin with.

Here's some more data showing an announcement for
150.182.208.0/20 originated by 26347

http://www.ris.ripe.net/mt/rissearch-result.html?aspref=150.182.208.0%2F20&preftype=EMATCH&rrc_id=1000&peer=ALL&startday=20130111&starthour=00&startmin=00&startsec=00&endday=20130111&endhour=19&endmin=16&endsec=26&outype=html&submit=Search&.submit=type

I can send you more data if you need it.
Just contact me off-list.

Cheers,
Andree