Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


I confused about OneProvider "2.5G" and "1G" difference...
New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

I confused about OneProvider "2.5G" and "1G" difference...

I got OneProvider Amsterdam Atom C2350 1Gbps(€7.99) and 2.5Gbps(€11.00).

I think "2.5G" is much faster than "1G".

But when I tested them... I found out ... not true at all...


1G

2.5G


1G

2.5G


Is this normal?

Comments

  • deankdeank Member, Troll

    Yes, it's normal.

    Thanked by 1eol
  • NeoonNeoon Community Contributor, Veteran

    Yes, if you think you can get fully 2.5Gbit for 10EUR, no.

  • Good luck with your budget.

    Thanked by 1Janevski
  • TheRealDealTheRealDeal Member
    edited February 2019

    Try this command and convert the MB/s into MBPS. It will give you a more accurate measurement because these test files are on a 10GBPS server.

    wget http://ovh.net/files/10Gio.dat -O /dev/null

    Thanked by 2eol xiaopigu
  • MasonRMasonR Community Contributor

    It's a marketing gimmick, check this thread for more details and lunacy:
    https://www.lowendtalk.com/discussion/147494/oneprovider-not-so-good-experience

    Thanked by 4eol imok xiaopigu masteri
  • Thank you all, guys.
    Now I know... I'm too stupid.

    Thanked by 2eol netomx
  • Daniel15Daniel15 Veteran
    edited February 2019

    If you're on Linux, make sure you tune the TCP stack, as the default settings are not sufficient for such a fast connection:

    In particular, the maximum receive and send buffers (net.core.rmem_max and net.core.wmem_max) need to be increased quite a bit, otherwise the buffer won't be large enough to keep up with the connection speed. For 2.5 Gb/s, I'd set them both to at least 32 MB (33554432 bytes). The auto-tuning parameters (net.ipv4.tcp_rmem and net.ipv4.tcp_wmem) need some adjustment too, but look at a guide to work out what to set them to.

  • MasonRMasonR Community Contributor
    edited February 2019

    @xiaopigu said:
    Thank you all, guys.
    Now I know... I'm too stupid.

    Not your fault that OneProvider has shady business practices. I'm sure they've trapped many customers with their misleading/false advertising.

    @Daniel15 said: ...

    OneProvider staff have confirmed in the past that 2.5Gbps speeds aren't achieveable on these servers due to the NIC and switch limitations. Check out the couple of OP responses in the thread I linked above. They're pretty much just doing what the car industry does -- "buy this awesome car that has a top speed of over 200 mph!" -- meanwhile, there's an undisclosed speed limiter that prevents it from going over 120.

    Thanked by 1xiaopigu
  • NeoonNeoon Community Contributor, Veteran

    @MasonR said:
    OneProvider staff have confirmed in the past that 2.5Gbps speeds aren't achieveable on these servers due to the NIC and switch limitations. Check out the couple of OP responses in the thread I linked above. They're pretty much just doing what the car industry does -- "buy this awesome car that has a top speed of over 200 mph!" -- meanwhile, there's an undisclosed speed limiter that prevents it from going over 120.

    They do sell servers with 2.5Gbit nics, depending on location.
    You can reach up to 2.5Gbit, but for how long is questionable.

    But true, its mostly french canadian clickbait.

    Thanked by 1MasonR
  • MasonR said: OneProvider staff have confirmed in the past that 2.5Gbps speeds aren't achieveable on these servers due to the NIC and switch limitations.

    Wooooow. I didn't see that. My comment was assuming that 2.5 Gb/s is actually attainable, but I guess it isn't. That's extremely misleading advertising - Why do they advertise one server as "1 Gb/s" and another as "2.5 Gb/s" if both can only reach 1 Gb/s maximum?!

    Thanked by 1MasonR
  • rm_rm_ IPv6 Advocate, Veteran
    edited February 2019

    Check with:

    iperf3 -R -c iperf.online.net -p 5201

    Change 5201 to higher numbers up to 5209 if it says "server is busy".

    Browser-based tests do not work well for very high speeds, and also you shouldn't expect 2.5 Gbit (or even 1 Gbit most of the time) to anywhere outside of Online's own network.

    Also, since your server is in Amsterdam, even the test I proposed above might be affected by them still fighting the recent fiber cut issues: http://netmap.scaleway.com/
    The iperf server is in Paris, and links from Paris to Amsterdam are heavily loaded, which may affect peak speeds.

  • i had one of these when they were sold by online.net - it only means they are connected to the switch via an 2.5gbit optical link. but you are not the only person on that switch and share it with a butt load of people.

    i could actually boost some torrents to full speed, but that was when these things were new, i bet they made some adjustments since then. also it's hard to achieve with their shitty HDDs, as soon the RAM buffer is fully used, the speed will drop sub gbit

  • Doesn’t every other VPS provider do this too?

  • @corbpie said:
    Doesn’t every other VPS provider do this too?

    Pretty much. Some don't lie about the hardware, though.

  • That's why call it 2.5G fair usage

  • @Etnous said:
    That's why call it 2.5G fair usage

    I think it's not even 2.5G to begin with.

  • @eol said:

    @Etnous said:
    That's why call it 2.5G fair usage

    I think it's not even 2.5G to begin with.

    You think about a lot of things proven wrong though. fyi

  • Ur definitely not stupid.

    Just small butt.

    Thanked by 1xiaopigu
  • @smile said:

    @eol said:

    @Etnous said:
    That's why call it 2.5G fair usage

    I think it's not even 2.5G to begin with.

    You think about a lot of things proven wrong though. fyi

    Well, no one has been on-site to confirm or deny.

  • @eol said:

    @smile said:

    @eol said:

    @Etnous said:
    That's why call it 2.5G fair usage

    I think it's not even 2.5G to begin with.

    You think about a lot of things proven wrong though. fyi

    Well, no one has been on-site to confirm or deny.

    @Clouvider would know!

  • @hucken said:
    i had one of these when they were sold by online.net - it only means they are connected to the switch via an 2.5gbit optical link. but you are not the only person on that switch and share it with a butt load of people.

    i could actually boost some torrents to full speed, but that was when these things were new, i bet they made some adjustments since then. also it's hard to achieve with their shitty HDDs, as soon the RAM buffer is fully used, the speed will drop sub gbit

    both 2.5G


    1G

    2.5G

    I know i am not the only one person on that switch.

    I know i can't use full speed entire time.

    but the thing is... 2.5G price is higher than 1G... so the speed should be faster...

    Thanked by 2Daniel15 eol
  • That's just the NIC though.

  • NeoonNeoon Community Contributor, Veteran

    @eol said:

    @Etnous said:
    That's why call it 2.5G fair usage

    I think it's not even 2.5G to begin with.

    It depends on the DC, they got 2.5Gbit.
    But its the same with Hetzner, dedicated until you reach the switch.

    Then its shared, har.

  • My Hetzner dedi still reaches 100MB/s in down and upload, even during Peak times.

    Thanked by 1eol
  • Hetzner sustained 1Gbit no problem if the other side can do it.

    Thanked by 1Hetzner_OL
  • @TheRealDeal said:
    Try this command and convert the MB/s into MBPS. It will give you a more accurate measurement because these test files are on a 10GBPS server.

    wget http://ovh.net/files/10Gio.dat -O /dev/null

    MB/s is the exact same as MBPS. You probably meant MBytes/s to MBits/s.

  • NeoonNeoon Community Contributor, Veteran

    @pike said:
    My Hetzner dedi still reaches 100MB/s in down and upload, even during Peak times.

    Well, some people complained about getting only 300Mbit, the question is just to where.
    I guess they did not checked the local network.

    Thanked by 1pike
  • @TimboJones said:

    @TheRealDeal said:
    Try this command and convert the MB/s into MBPS. It will give you a more accurate measurement because these test files are on a 10GBPS server.

    wget http://ovh.net/files/10Gio.dat -O /dev/null

    MB/s is the exact same as MBPS. You probably meant MBytes/s to MBits/s.

    Indeed I did mean Mbytes/s to Mbits/s this is the one time autocorrect got me.

    Thanked by 1TimboJones
Sign In or Register to comment.