[FFmpeg-devel] FFmpeg/MPlayer/rtmpdump possibly searching for a new server and hosting

Timothy Gu timothygu99 at gmail.com
Wed Jul 15 10:00:48 CEST 2015


On 7/15/15, Michael Niedermayer <michael at niedermayer.cc> wrote:
> On Tue, Jul 14, 2015 at 11:42:30PM +0100, Kieran Kunhya wrote:
>> OVH has poor quality connectivity by the way - this could lead to
>> performance issues in some countries.
>
> Has anyone done some hoster connectivity comparissions /tests ?
> it would be interresting to see what the differences are.

Trying to download from OVH (http://bhs.proof.ovh.net/files/100Mio.dat)

- In China's home network (Shanghai): avg. 9 KB/s
- In DigitalOcean's LAX1 (Los Angeles) data center: avg. 11 KB/s
- In DigitalOcean's NYC3 (New York) data center: over 10 MB/s
- In AWS EC2 us-west-2 (Oregon, US): over 2 MB/s
- From fate.ffmpeg.org (somewhere in France): about 13 MB/s

In comparison: DigitalOcean FRA1 (Frankfurt I believe)
(http://speedtest-fra1.digitalocean.com/100mb.test)

- China: avg. 2 KB/s (yeah, that's right)
- DO LAX1: avg. 10 MB/s
- DO NYC3: avg. 10 MB/s
- AWS us-west-2: avg. 2 MB/s
- fate: 100 MB/s

In comparison: AWS EC2 eu-central-1 (curl -o /dev/null
http://eu-central-1.ec2.cloudharmony.net/probe/test304kb.jpg)

- China avg. 5 KB/s
- DO LAX1: avg. 10 KB/s
- DO NYC3: >400 KB/s (can't be measured accurately)
- AWS us-west-2: >100 KB/s (can't be measured accurately)
- fate: >2 MB/s (can't be measured accurately)

Only the China test is actually from a home connection. Everything
else is from a dedicated server connection.

HTH.

Timothy


More information about the ffmpeg-devel mailing list