[6bone] [NOTIFY] XS26 service/peering outage

[Petr Baudis]

  Hello,

  we want to notify you about a XS26 service and peering outage, which is
expected to occur in the next days, starting by Dec 23 2002.

  Due to consistent and major problems with zebra as a routing software
(BGP+OSPF6) on our Points of Presence and fundamental lack of certain features
we require in the routing protocols we currently use, we decided to switch the
whole network to a new completely different routing software, coded by Jan
Oravec specially for the Access to Six project. This software will use own
special protocol for internal routing inside of the XS26 network and it will
support BGP for external peerings. This software will also make it possible to
finally add support for user BGP peerings and dial-up users support. Also,
considerable web interface usability improvements will be done.

  The current XS26 network is de facto unmaintained for a few last days. Zebra
didn't scale well at all and the frequency of outages and crashes of its
instances on various PoPs was exceeding the critical values lately. Note that
we are not aware of any problems caused to our eBGP peers because of these
issues.

  We are fully aware of the current routing problems and outages, thus we
decided to introduce the routing software as soon as possible. However, the BGP
support isn't yet production-grade and we are afraid of messing up the global
BGP tables, causing pan-6bone damages. Thus, we decided to keep the old bgpd
running on nlams-02-01 and dkcop-0a-00 with BGP sessions to AORTA (NL and NO),
advertising only ourselves and making AORTA (formerly known as CHELLO) our
default route. Thus, for this period (we expect the full BGP support to be a
matter of days), our prefix will be announced via AORTA and it will be our only
connectivity. This way, we would also like to thank AORTA for the support and
great help.

   Even after starting up our BGP implementation we will announce only our
prefix (we won't provide transit) for some time, while we will be testing it
carefully. We expect to proceed with re-establishment of the BGP peerings
slowly and carefully, as we don't want to harm connectivity of other sites or
pollute the global BGP table with bogus entries. Also, we will take this
opportunity and give full-transit only to those who will actually want it
explicitly.

  If you would notice any such behaviour, please notify us in case we would not
notice that immediately. We hope that the transition problems will be minimal,
but it's expectable that since the software was not tested on any other
production network yet, there will be maybe some problems and further outages.
Please bear with us, we will try to correct any problems as soon as possible.

  Kind regards,

			Petr Baudis <pasky@xs26.net>
			(on behalf of the XS26 team)