[Andy]

Are there any future developments to get peering?

[Neil]

Hi

We do not have peering directly to Sweden so it does rely on Transit providers, since this forum is not an official support method the best thing to do is send us an email with ping and traceroute from and to the server to the Swedish IP, then we can check and see what we can do.

[Andy]

Some reply from OVH would be nice. I love how you always ignore problems...

[Thelen]

This is in part why they have increased the price of extra BW, and why they are switching to the new per-account stuff shortly.

You can't compare 10Gbit on i3d to OVH, even with the lack of performance currently...

[turbanator]

thats bad considering they are not guranteeing the 10gbit with HG servers. They have a huge set up fee..for around 400 usd u can easily get 10gbit servers from i3d dedicated bandwidth with 100tb out with top notch hardware and network. They should not be doing that to HG range..MG range understandable but not for HG..the set up fees are too big. Also yes serious upgrades are needed.

[virtuallynathan]

Their network is very clearly congested... Peak time I can manage 60MB/s from proof.ovh.net, non peak i've seen up to 600MB/s. I have mentioned these issues to Octave/Oles and his response was that they are no longer guaranteeing 10Gbit/s on HG servers. You can see from the weathermap that more upgrades are needed in the EU and US.

[turbanator]

amen..it is very frustrating...

[Andy]

Yet another slow speed issue to Sweden (what's wrong with these connections all the time?).

A friend asked me to test the download speed from his server hosted at his university. I managed a solid 10MB/s from some Linux ISO's hosted on it, as he expected as it's on a 100Mbps switch. However when I asked him to download from me the other way, he only managed 250KB/s!!!

His ISO's are here: https://stormhub.org/linux/ so you can see that the speed is as expected.

Here is a tracert from my server to his server:

1 <1 ms <1 ms <1 ms rbx-51-m2.fr.eu [91.121.167.252]
2 <1 ms * <1 ms rbx-2-6k.fr.eu [91.121.130.130]
3 <1 ms 2 ms 2 ms rbx-g2-a9.fr.eu [91.121.131.10]
4 5 ms * * ams-5-6k.nl.eu [91.121.131.174]
5 5 ms 5 ms 5 ms nl-sar.nordu.net [195.69.145.203]
6 18 ms 20 ms 18 ms dk-uni.nordu.net [109.105.97.26]
7 34 ms 34 ms 34 ms se-tug.nordu.net [109.105.97.9]
8 34 ms 34 ms 34 ms t1tug.sunet.se [109.105.102.18]
9 34 ms 74 ms 119 ms t1fre-ae0-v1.sunet.se [130.242.83.37]
10 61 ms 34 ms 34 ms m1fre-ae1-v1.sunet.se [130.242.83.45]
11 34 ms 34 ms 34 ms ls-kth-br1.sunet.se [193.11.0.194]
12 34 ms 34 ms 34 ms cn6-br1g-p2p.gw.kth.se [130.237.0.2]
13 37 ms 36 ms 44 ms vss1-cn6-p2p.gw.kth.se [130.237.211.117]
14 * * * Request timed out.
15 * * * Request timed out.
16 * * * Request timed out.

His server to mine:

1 router2.math.kth.se (130.237.41.2) 1.402 ms 1.386 ms 1.347 ms
2 cn6-vss1-p2p.gw.kth.se (130.237.211.118) 0.695 ms 0.885 ms 1.146 ms
3 br1g-cn6-p2p.gw.kth.se (130.237.0.1) 0.387 ms 0.367 ms 0.344 ms
4 m1fre.sunet.se (193.11.0.193) 0.376 ms 0.558 ms 0.532 ms
5 t1fre-ae5-v1.sunet.se (130.242.83.46) 0.510 ms 0.515 ms 0.746 ms
6 se-fre.nordu.net (109.105.102.9) 34.539 ms 34.450 ms 34.438 ms
7 dk-ore.nordu.net (109.105.97.6) 9.771 ms 9.770 ms 9.782 ms
8 uk-hex.nordu.net (109.105.97.29) 30.757 ms 30.746 ms 30.718 ms
9 * * *
10 rbx-g2-a9.fr.eu (91.121.128.195) 35.358 ms 35.347 ms 36.747 ms
11 rbx-2-6k.fr.eu (91.121.131.9) 34.961 ms rbx-1-6k.fr.eu (91.121.131.13) 35.909 ms *
12 rbx-51-m1.fr.eu (91.121.130.25) 35.009 ms 34.961 ms rbx-51-m1.fr.eu (91.121.130.153) 34.908 ms
13 * * *
14 * * *
15 * * *

Why is OVH relying on it's customers to inform you of speed problems? Surely you have some sort of process in place to ensure speed and latency is at it's best all of the time and that links are not saturated? Yet these problems still crop up?

I pay for an OVH.co.uk server, £84 a month, I expect more than 250KB/s when the other way around I can get 40x that. I'd like answers please.

Edit: My friend now tells me he can get 7MB/s spikes on occasion but it's sporadic. His connection isn't overloaded, it's a 10Gbps connection for the university servers only, and regardless the downstream should be less saturated than the upstream.