 |
Linode Forum Linode Community Forums
|
| Author |
Message |
tmslnz
Joined: 05 Nov 2010
Posts: 2
|
| Posted: Sat Nov 06, 2010 9:24 am Post subject: Cheeky test Linode 512 vs RackSpace 256 |
|
|
Alas, I don't think the ram difference counts.
The two machines are set up with the same packages, the Linode one is also actually serving stuff, while the RackSPace one is idling.
Click for video: (Linode on the left)
Speed.
But more seriously, is it mostly network latency? My Linode is in the UK and I am in London.
EDIT
In case you are wondering, what I am doing is a straight MySQL dump via Sequel Pro on OSX, connected via SSH. |
|
| Back to top |
|
hoopycat
Joined: 30 Aug 2008
Posts: 1294
Location: Rochester, New York
|
| Posted: Sat Nov 06, 2010 10:55 am Post subject: |
|
|
Network latency may well play a part, but a single long TCP session will usually adapt to that quickly. If there is a lot of back-and-forth query-and-response activity, though, you'll run into the laws of network physics on each exchange. (See this forum thread for my previous ramblings on this topic.)
There's always a bottleneck somewhere (disk I/O, network I/O, CPU, etc), and figuring out what it is (or what they are) would be excellent in both cases. Right now, there's too many variables to extend this beyond "Linode is probably faster at exporting databases to London people who have a $20/mo budget" :-)
(This post counts as studying for my data analysis exam.) |
|
| Back to top |
|
tmslnz
Joined: 05 Nov 2010
Posts: 2
|
| Posted: Sat Nov 06, 2010 11:48 am Post subject: |
|
|
| True, true, but still. My clients are in the UK, I am in the UK, and performance seems to heavily lean in favour of Linode throughout, not just at exporting MyQSL. Good stuff, but I'm not a computer scientist. |
|
| Back to top |
|
| |
|