![[H]ard|Forum](/styles/hardforum/xenforo/logo_dark.png){kind=logo}
agrikk
Gawd
- Joined
- Apr 16, 2002
- Messages
- 933
I've recently become an employee of Amazon Web Services, and one of the perks is that I get a free AWS account that essentially allows me to run what I want wherever I want.
I haven't bothered to turn up a monster EC2 instance to fold on it, but in order to learn a bunch of their different product offerings, I'm in the process of building a F@H stats site here:
http://folding.rainwalk.net
As you will quickly see, the site is retardedly rudimentary but I'll be adding functionality to it over the coming months. I can't imagine it ever approaching the usability of the Extreme Overclocking site, but I plan on building on this nonetheless. Feel free to give it a look, recommend features or ask questions about AWS' services.
The main thing that differentiates my data from the EOC Folding site is that I don't prune user or team data and I update every hours instead of the 90 minute cycle.
The EOC database contains 6000+ teams and 500k+ users, whereas my database includes the full 89,000 team data and the 1.5 million users in the F@H database. I've had a database running in my colo for almost a year now, just basically sitting there collecting hourly snapshots, and I figured it was time to do something about it.
The site is driven by four web servers split into two pairs, one each on the east coast and west coast that sit behind load balancers and connected via Route 53 DNS services to provide geographical latency-based load balancing. The back end is powered by a pair of database servers, one each on the west coast and east coast that receive data from a utility box that pulls data from F@H, parses it and shoves it out to the two database servers.
For kicks, enter your user name in the search box and tell me which web server you landed on:
If you want to help even further, click on your name and tell me how long it takes to load your daily summary page:
I haven't been able to load test this setup, so I have only a minimal idea how it will perform under random user load. But right now it runs great with only me hitting it!
I haven't bothered to turn up a monster EC2 instance to fold on it, but in order to learn a bunch of their different product offerings, I'm in the process of building a F@H stats site here:
http://folding.rainwalk.net
As you will quickly see, the site is retardedly rudimentary but I'll be adding functionality to it over the coming months. I can't imagine it ever approaching the usability of the Extreme Overclocking site, but I plan on building on this nonetheless. Feel free to give it a look, recommend features or ask questions about AWS' services.
The main thing that differentiates my data from the EOC Folding site is that I don't prune user or team data and I update every hours instead of the 90 minute cycle.
The EOC database contains 6000+ teams and 500k+ users, whereas my database includes the full 89,000 team data and the 1.5 million users in the F@H database. I've had a database running in my colo for almost a year now, just basically sitting there collecting hourly snapshots, and I figured it was time to do something about it.
The site is driven by four web servers split into two pairs, one each on the east coast and west coast that sit behind load balancers and connected via Route 53 DNS services to provide geographical latency-based load balancing. The back end is powered by a pair of database servers, one each on the west coast and east coast that receive data from a utility box that pulls data from F@H, parses it and shoves it out to the two database servers.
For kicks, enter your user name in the search box and tell me which web server you landed on:
If you want to help even further, click on your name and tell me how long it takes to load your daily summary page:
I haven't been able to load test this setup, so I have only a minimal idea how it will perform under random user load. But right now it runs great with only me hitting it!
