OVH Community, your new community space.

Roubaix <> London - 09/01/2010 Questions


tjm21
16-12-2009, 14:37
It's weird I have one server with a 91. ip address and can max out my VM 50mb broadband and one with a 94. ip address and rarely get over 2MB. this is even downloading the same file with between 10 to 20 threads.

IainK
16-12-2009, 14:10
Lucky for some :P I'm having troubles pulling 2.0MB/s from any OVH servers over vmedia even with 8-16 threads.

freshwire
16-12-2009, 00:11
All is good with virgin media. Combined (2 threads) speed of 5.4 MiB/s using filezilla. My line is 50 Mb/s.


Andy.
15-12-2009, 23:56
This is the first time ever I've had customer complaints from UK users about slow FTP on a OVH server.
Thing is, other UK customers on other data center servers we use dont have this issue, so yeah, whatever, something is going on and a burnt cable that needs replacing but isn't replaced for weeks may or may not have something to do with it so just make up your own story as we'll get told zip here.

IainK
10-12-2009, 20:09
Bloody hell. That is unbelievable that LINX has only '2 peering platforms' and a whole 1 was crawling to a hault yesterday. For a network that covers almost all UK ISPs they need to get that stuff sorted.

Andy
10-12-2009, 18:22
Quote Originally Posted by adama
That's just LINX's graphing being retarded.

If LINX had dropped 90% of their traffic today, believe me, it would have been on the News.

Lost access to my server now. Can't reach it from another OVH machine either. :<

adam.
Err, it was on the news

http://www.theregister.co.uk/2009/12/10/linx_failure/

Myatu
10-12-2009, 11:35
LOL! You're only off by 10 days...

gigabit
10-12-2009, 11:33
That's why I'm confused, I thought we were still in November.

Myatu
10-12-2009, 11:32
They borked vss2 last night/this morning (depending on your schedule ). See http://forum.ovh.co.uk/showthread.php?t=3202

gigabit
10-12-2009, 11:27
Me too, getting lots of "server down" texts this morning between 3 and 4AM... What the hell is going on?

IainK
10-12-2009, 11:06
My business partner reports our servers being unavailable after 3am.

Euan
10-12-2009, 10:54
Quote Originally Posted by adama
That's just LINX's graphing being retarded.

If LINX had dropped 90% of their traffic today, believe me, it would have been on the News.

Lost access to my server now. Can't reach it from another OVH machine either. :<

adam.
It was pretty bad for most of us traffic was being dropped from around 8pm to midnight GMT.

adama
10-12-2009, 04:45
That's just LINX's graphing being retarded.

If LINX had dropped 90% of their traffic today, believe me, it would have been on the News.

Lost access to my server now. Can't reach it from another OVH machine either. :<

adam.

NickW
10-12-2009, 02:01
Yeah, LINX look like they've had some issues. https://www.linx.net/pubtools/traffi...html?stats=day

mattb37
10-12-2009, 00:03
Would I be correct saying OVH use LINX for peering in London?

I have had an email from another data centre where I have servers reporting packet loss and a second provider also reporting reporting packet loss in London, both use LINX. One provider has said it is a problem within the LINX infrastructure. This could explain slow loading, correct?

gigabit
09-12-2009, 21:53
People in UK have reported my site not loading/slow, whereas US people saying its slick as usual. *shrug*

Myatu
09-12-2009, 21:42
Quote Originally Posted by Andy
I mis-read it then, but even so there has been issues, lol. All I saw was Roubaix <> London, fibre and cut
No, you read it right. "cable damaged by a fire" -- Unless they're planning a controlled fire for the 9th of January!

I do have some connectivity issues though. I'm running over a VPN right now so it helps a little since its over UDP, which is a bit more forgiving than TCPs outright disconnects.

PS: A traceroute to your server goes through vss-2-6k.routers.chtix.eu, which is glowing red in the weathermap.

NickW
09-12-2009, 21:32
Yeah, it seems to be fixed now though :P.

On another note, upgrade VSS1 and 2 OVH. Massively in the red on weathermap.

Andy
09-12-2009, 21:30
I mis-read it then, but even so there has been issues, lol. All I saw was Roubaix <> London, fibre and cut

gigabit
09-12-2009, 21:28
Fun fun fun...

NickW
09-12-2009, 21:26
This outage will be in 2010.

The current issue seems to be with a switch in London sw1.tc.lon.ovh.net me trhings from pings/traceroutes. Currently having to connect to this forum via my US server - it traceroutes from New York straight to Paris so I avoid London, as it wouldn't load for me otherwise.

Andy
09-12-2009, 21:22
Anyone else getting extreme packet loss and pathetic speeds due to the re-routed traffic? Its pathetic. I'm trying to remote desktop in to set up IRC and services, and I can't because it keeps disconnecting.

Myatu
09-12-2009, 19:13
Hah! Felt like traffic was flowing through Timbuktu last night (1 am-ish). The weathermap for London was - and still is at times - glowing red on vss-1 and vss-2, with the rest pretty much approaching red. RBX2 weathermap has a nice Christmas feel as well, with lots of red on a white background

PS: What fire? Aren't there fire suppressors to should have avoided this? Or was it outside the DC?

Neil
09-12-2009, 12:51
Hi

All the traffic will be rerouted, so latency, pings and speeds maybe affected. Most of the traffic will go via Amsterdam or Paris, but it depends on your ISP, so don't blame us if it goes via Timbuktu ;-)

Paw-Fox
09-12-2009, 11:49
I would imagine an backup route would come up. - And no ISP's shouldn't break at all.

MicroChip123
09-12-2009, 11:39
I was reading http://travaux.ovh.com/?do=details&id=3639

Our fiber Roubaix <> London will be shut at night from 09 to 10/01/2010 due to repairs (replacing a portion of cable damaged by a fire). The cut will take place from 00:00 CET for a maximum of 6 hours.
Does this mean all UK IP's will be broken during that time or will they be routed via other fibers?