21–30 of 86 posts
Something doesn't add up here. They're actively telling third parties to lease cloud servers with large amounts of free bandwidth, and allocate them to this purpose. I see four possibilities: - They are paying the third party less than what it costs to rent the server, in which case the third party is losing money and shouldn't be doing this. - They are paying the third party at least what it costs to rent the server…
Something doesn't add up here. They're actively telling third parties to lease cloud servers with large amounts of free bandwidth, and allocate them to this purpose. I see four possibilities: - They are paying the third party less than what it costs to rent the server, in which case the third party is losing money and shouldn't be doing this. - They are paying the third party at least what it costs to rent the server…
So, if this becomes popular, flat rate servers with unmetered bandwidth at a fixed speed would become more expensive, or come with a more specific ToS to prohibit that.
Similar to how data hoarders ended the "really cheap personal cloud storage" bubble.
Something doesn't add up here. They're actively telling third parties to lease cloud servers with large amounts of free bandwidth, and allocate them to this purpose. I see four possibilities: - They are paying the third party less than what it costs to rent the server, in which case the third party is losing money and shouldn't be doing this. - They are paying the third party at least what it costs to rent the server…
The more this trick will be used, the higher the cost of the dedicated fast internet connections with cloud hosts will become. Most people buying dedicated, unlimited uplinks don't upload the amounts mentioned in this article, which is why bandwidth is relatively affordable. If everyone buys into this network to make a profit, the prices would probably rise to the point that profit drops. This isn't against any kind of ToS I know of.
The "make 2.4x your investment" promise also assumes your server will actually serve 100TB of content. The more people join, the fewer files your server will serve and the less profit you'll make.
In the end, the system is profitable until the network has reached the size where everyone breaks even. If people leave the network, profits go up again, if people join the network, profits drop and people back out. People renting servers game the system trying to one-up each other and in return Farba gets a worldwide CDN for cheap.
The best trick may perhaps be that server providers may directly buy into the system, as they can get bandwidth and servers for lower margins than their customers. This gets Farba high-quality service for cheap while the service providers add a relatively easy income stream on the side.
All of this comes at an additional cost to the customers, because now your battery-starved mobile phone is validating cryptographic signatures of images and other content.
This also leaves legal questions: who's responsible for illegal content hosted on servers hooked into this network? If your hosting provider of choice gets a takedown notice for child porn or malware, will you still be able to use their services?
There's a bunch of really obvious objections to this idea around performance, privacy, economics, UX, and reliability. Impressively, the post did not manage to address any of them up front. 1. Performance. In terms of latency, this is adding an extra round-trip to each page request (to get the metadata). It's also likely going to load different resources from different servers (hostname, IP) preventing connection reu…
>1. Performance. In terms of latency, this is adding an extra round-trip to each page request (to get the metadata). It's also likely going to load different resources from different servers (hostname, IP) preventing connection reuse. I guess they can't even start fetching data until the full page has loaded? fair point. We use same technique as lazy-load (you have to wait the full page has loaded). >3. Economics. If…
1. It started out burning huge amounts of investor capital, basically transferring all of that money to drivers and in the form of subsidized rides, to riders.
2. It works today by a combination of the above and underpaying drivers, who generally make less than minimum wage when they manage to cover their completely costs at all. Despite this, prices have gone up dramatically, so that taxis are once again competitive on a price basis.
3. Uber was profitable in 2018, and not before nor since[0]. Deeply, profoundly unprofitable. Why would anyone want to be the Uber of anything?
0. https://www.macrotrends.net/stocks/charts/UBER/uber-technolo...
However, a lot of people use CDNs because they are fast. They want the latency to be ultra low and they want it to be reliable. This is where the author's solution falls short: CDNs can get premium quality bandwidth because they have more scale than their customers. Sure you can find servers from various hosts in lots of different data centers, but you might not get the latency or quality of bandwidth you were looking for. This can be critical for some applications like streaming or serving images.
If your constraint is speed, you almost have to go with a CDN (even if you are huge). If it's volume, you are in a better position. If your volume is small, you might as well go with something off the shelf like Cloudfront or Cloudflare. If you are ultra huge like YouTube or Netflix, then you have to peer directly with ISPs.
There are several red flags here, not the least of which is that a CDN isn't just about bandwidth, it is about disk . If a cache node doesn't have the object, it has to go to origin to get it. That has a cost at origin, which most CDN customers want to minimize. Cache management (eviction, invalidation, etc) is done where? By the look of that diagram, you are actually doubling your origin traffic just to push to the…
The disk performance - is the weakest part of the Peer. It is a part you can't guarantee any QoS on a cheap server. The idea is to keep popular cache on Peers in memory. Everything else we will host from a limited number of regular servers (belong to Farba), with high-performance disk subsystem.
>Cache management (eviction, invalidation, etc) is done where?
yes. You can invalidate either a single URL or a folder with a wildcard
>you are actually doubling your origin traffic just to push to the cache after serving the original client request
this is something we can proxy and do only one request per file. Current implementation was done for simplicity and reliability.
>but I don't see how this can work. Are they going to rewrite the object in a way that doesn't require the host certificate?
you can easily check this up. Please open the network console on our demo-page and see how it works:
- we have two different certificates (one for the main site, for for the Peer)
- everything works flawlessly
There's a bunch of really obvious objections to this idea around performance, privacy, economics, UX, and reliability. Impressively, the post did not manage to address any of them up front. 1. Performance. In terms of latency, this is adding an extra round-trip to each page request (to get the metadata). It's also likely going to load different resources from different servers (hostname, IP) preventing connection reu…
>1. Performance. In terms of latency, this is adding an extra round-trip to each page request (to get the metadata). It's also likely going to load different resources from different servers (hostname, IP) preventing connection reuse. I guess they can't even start fetching data until the full page has loaded? fair point. We use same technique as lazy-load (you have to wait the full page has loaded). >3. Economics. If…
Earlier quoted context omitted.
>1. Performance. In terms of latency, this is adding an extra round-trip to each page request (to get the metadata). It's also likely going to load different resources from different servers (hostname, IP) preventing connection reuse. I guess they can't even start fetching data until the full page has loaded? fair point. We use same technique as lazy-load (you have to wait the full page has loaded). >3. Economics. If…
Completely skipping over the privacy concern doesn't seem like an amazing look
Zero-cost CDN for 2 USD/TB! Sounds like a great deal to me.
On a more serious note, is price the problem with other CDNs?