[Mageia-sysadm] is the svn server down?

[Michael Scherer]

Le vendredi 30 mars 2012 à 22:21 +0200, nicolas vigier a écrit :
> There is a problem with the switch where our servers are connected.
> We may need to replace it.

mtr ryu and mtr alamut show the problem is just at the entry of the ielo
network :

on ryu, there is 6 hops after 
xe2-2-0.ter2.eqx2.par.as8218.eu

ge1-5.br1.rdb.par.ielo.net
ge-1-6-neo-cr2.eqx-prs.fr.rt.ielo.net
2ge-e1-17-e1-18-cr5.le9-lyon.fr.rt.ielo.net
2ge-e1-17-e1-18-cr4.le9-mrs.fr.rt.ielo.net
ge-2-1-ve25-cr3.ldr-mrs.fr.rt.ielo.net

I am not a expert into deciphering network admin language, but I suppose
that :
- since that's all ielo.net, that's also the same type of convention
used for all equipements.

ge usually mean "gigabit ethernet", 2ge maybe is when there is 2 cards,
or something like that.

ldr-mrs mean "liandier marseile" ( ie, the location of our hoster,
liandier street, in Marseille ), so that's a network device in the
server room.

The 2 others one are gigabit switch , 1 in Marseille, the other one in
Lyon. le9 is the french for "the 9", the old name of a ISP now merged
with SFR ( another french isp/telephony provider ). If we see the list
of PoP of ielo ( a pop being a equipment where someone can connect ) :
http://www.ielo.net/solutions/infrastructures/points-de-presence/ 

We see that indeed, they are in Lyon and Marseille in "le neuf"
datacenter ( ie, le 9 ).

There, they are connected to lyonIX, as seen here 
http://www.lyonix.net/fr/a-propos-de-lyonix/participants-lyonix
( a IX being a place where ISP interconnect their network, seek
"internet exchange" on wikipedia )


- ge-1-6-neo-cr2.eqx-prs.fr.rt.ielo.net

and then, they go to the equinix datacenter in Paris ( eqx-prs ), and
they are connected to as8218.eu, that is neotelecom ( hence "neo" ). 

See for the detail of the as 
http://www.peeringdb.com/view.php?asn=8218

( a AS being a autonomous system, think of it as a network under the
control of one single administrative entity, the internet being a
network of network, ie interconnection of AS ).


for the mageia.org servers, there is nothing. 

At first sight, I would say that having our switch ( ie, the small
consumer grade equipment we use to interconnect to the main
cisco/hp/whatever switch of Lost oasis ) would not cause such issues
( ie, I would expect to drop paquet at the last step ). However, since
all others servers on the same Lost Oasis switch seems to work, that's
also weird.

For the record, I see that 212.85.158.145 ( alamut default gateway,
whose reverse is gateway-mageia.ielo.net ) is not answering to ping and
on the other hand, the default gateway for zarb.org ( 212.85.158.1 )
answer to ping. 

And I am pretty sure that this gateway is after our switch from a
network point of view, and so that this should answer fine if our switch
is unplugged.  

So I suppose that this device ( 158.45 ) do send network
topology/routing with OSPF or RIP etc, and for some reason, it no longer
announce the route for our network.

It could be caused by our switch being broken, but that would be rather
strange, as I do not know any network equipment that would act like this
( doesn't mean that it doesn't exist, but usually, network is not that
smart ).


So what did IELO said exactly ?

( cause "our switch" can mean the switch we use, or the switch they use
for us, as I doubt they remember do know exactly how everything is done
for all the client they have  )

And more important, what can we do to diagnose the issue, does it have
to wait on Monday ? 

-- 
Michael Scherer