SAFEcrypt Fiber and monetization of bandwidth on SAFE Exchange

#2

Right now… there’s a lot of ideas begging to explode. It’s like going to a new candy store and having so many choices… where to start???

I want to focus on the biggest demand “potential” with a large consumer base “already waiting”. It looks like PUT bandwith (uploading) will be in very high demand because of limited speed/access combine with potential demand after SAFE goes public. This is core to farming and consumer uploading.

I hate to split focus, but I do believe they can done simultaneously.

Farming Vault Demand
There are many farmers that don’t have “reliable high upload bandwith” needed to run a large farm. Their choices are likely…

Consumer Upload Demand
There may not be many at first, but eventually there will be people who want to upload their entire collection of… whatever, but have very slow upload speed.

At that point, I hope we develop a “grandma just button” way for them to do it. It’s like those express lanes people use to get to their destination faster. People are always willing to pay more for convenience.

Bringing it all together
Assuming SAFE Network PUT storage becomes as cheap as we hope…
Assuming upload bandwith gains demand from farming and consumer uploading…
A co-op providing premium upload bandwith service will be in a great position to meet both demands and help mass adoption.

I would like to talk in more detail about the following.

  • Developing SAFEcrypt Drive delivery system.
  • Integrating payment system to accept all crypto currencies.
  • Co-op funding after SAFE goes public.

Depending on the discussion, this may end up being my main focus. It involves farming, and can tap into future upload demand.

Here’s the bottom line

  1. The main revenue stream (Safecoin) comes from vault farming.
  2. Additional revenue stream (alt coins) comes from uploading service.
  3. The total budget from the Co-op will determine the starting scale of the operation. But unlike a crowd sale which is a one-time deal, the Co-op allows others to join later, adding more funds to the pool, similar to mining pools, except the farmers are “silent,” meaning they are not actively maintaining the vaults.
6 Likes
#3

You are speaking my language for sure. David I agree with your assessment exactly. I love how you explain your thought process…words and structure. It is rare to find things that have massive revenue potential and also a win for everyone. Having a focused group early with the intellect, variable skills, and resources to help ensure SAFE network takes off in the right direction seems like a great thing to be a part of.

The only graph I have on safecryptdrive.com shows the annual maintenance costs for companies to store 50 TB of data on their cloud servers. They also charge for put and gets and if the company stops paying then they lose all of the data and the put costs already spent. I won’t even get into the other revenue they get from controlling everyone’s data. The SAFEnetwork is a no brainer for companies needing dependable backup of their data. If there are not alternatives for people to provide high level bandwidth, then the few who control the bandwidth market now will make the obvious move and essentially stop SAFE in its tracks through bringing its speed to a stand still (Why do you think there is also a huge difference between consumer upload vs. download). We must decentralize the heck out of the big on/off-ramps to the internet network to give people alternatives that still offer a benefit over the current system, which SAFEcrypt Drive shows how easy it will be to get people interested.

Mesh is the obvious answer for consumer infrastructure, but development of the software and even more so the hardware will require massive investment. Having the ability to ship/mail large data to SAFE parts of the world that can quickly plug that data into the network will ensure the utility of the SAFE network.

I also see your point with cooperative-crowd fund, which is why I have started developing http://CrowdSAFE.io to be one of main places to launch these essential projects by connecting entrepreneurs/visionary’s with creative people and ensure a natural SAFE way of making sure all participants see an SAFE ROI. . 100-1000’s of Gbps bandwidth is available in my local area and we have the ability to become our own ISP with no third party between the global network and SAFE; although I have also found a much cheaper option of working with a local ISP that has already built the the network and is open to supporting a slower scaling timeline. Having the support (financial and talents) of thousands who all have an interest in SAFE being a success would be great to have protecting these few places in the world that have an open fiber network.

@ddabek has assured me @dyamanaka that there will be such crowdfunding-autonomous ROI distribution-voting feature on SafeX, which will be very valuable in focusing this talent and resources. It will be very nice when I can look on a price index and see the global price value of bandwidth and data storage (SAFE). All one has to do is look at what is IMHO being wasted on computational hash power for the Bitcoin network, purely for money. Imagine if those same creative/money-driven folks started getting a better ROI by providing new inventive ways to build a bandwidth infrastructure that connects the 4+ billion people in the world who do not have internet access and ensure the life of the SAFE network?

5 Likes
#4

Yes, this thread struck me to my core. It is so close to what I personally want to do regarding the SAFE farmer’s club. I’m sure others have been looking for this kind of opportunity too. I’m creating a new word “silent farming” which allows non-farmers to financially participate in our co-op for a return based on their percentage share of support every month. Here’s a rough idea how I see it working…

Upload Service + Silent Farming

A new user finds www.SAFEcryptDrive.com because they need to upload their home library. I think SafeX will be useful if they need to exchange their alt coins for SC to pay for the upload.

While they are on the site, they see another tab explaining how they can participate in silent farming. Those interested in earning SC will want to compare silent farming to alternatives.

  • $5 a month SAFEcryptDrive VS $5 a month Digital Ocean (Cloud Farming).
  • $5 yielded X SC, last month, this month, etc. (Silent Farming)
  • $5 buys X SC on the exchange. (Exchange Purchase)

Those (including myself) will keep adding funds every month as long as it yields more Safecoin per cost compared to alternatives. I’m hoping a bulk discount will keep the bandwith costs low enough to fund more vaults than I could on my own. If that is true then the only expense left is the hardware storage device, which is the 1TB USB keys already bought for the upload service… synergy!

The exchange price may be our biggest threat, because so many SC already exist from the crowd sale. Hopefully, the demand for SC will drive the fiat price beyond our costs and it won’t be a problem. Considering how much money is spent on blockchain mining, I’m not too worried. See solution below…

Additional Business Strategy
In World of Warcraft the gold price for the item I farmed sometimes dropped insanely low… I shifted gears and bought it off the auction instead of farming. @ddabek knows what I’m talking about. I think this should be part of the business strategy for the co-op.

  • If the SC exchange price < Silent farming… then funds go to purchasing SC off the exchange.
  • If the SC exchange price > Silent farming… then funds go to farming operations.
3 Likes
#5

@dyamanaka I’m smelling what you are stepping in…and I love it!:wink:

There are a lot of moving parts here and crowd sourcing this project is definitely the way to go. I will focus on web page design with creating a rough draft with the examples above; and also the things we discussed PM. If anyone out there has a talent for graphics to visually describe the above concept then that would be a huge help*anyone enjoy designing?

I will also search out the best source/price for 256/512/1000GB flash drives.

I will get exact monthly costs for bandwidth. Will start with 1Gbps upload/download and then also the next level up prices. I think I have all the hardware needed to start the upload service portion. hint having a hybrid farmer that streamlined uploading these flash drives and also farming would be cool :stuck_out_tongue_winking_eye: I currently have a consistent 15Mbps upload in my home right now. The fiber line is available of few blocks away so there are several options I will pursue…for bare minimum start up costs.

@ddabek any feedback with regard to contract-crowd supported project like this and leveraging it on SafeX, where we can take the human element out of the parts that include individual control over funds?
I am just stumbling onto DASH and I like their “Decentralized Governance by Blockchain” and “Two-Tier Network: Incentivized Full Node Operation” that use of 1000 DASH minimum account balance to run a masternode and have a stake in 45% return on transaction fees and voting rights of the 10% transaction fees towards development and 45% to the miners…link https://www.dash.org/i-understand-crypto/

We will also need a coder to take the current published SAFE libraries to create a UI and APP to encrypt/decrypt storage drives for the offline features of this project.

4 Likes
#6

Awesome @MrAnderson,

I think we had a good hangout tonight and molded the business concept into a more feasible yet flexible strategy. Duo purposing the (bandwith pipleline) and (USB storage devices) allows us to shift gears if necessary.

Our current road map is as follows…

Upload Service

  • We need a 1Gbps internet pipeline.
  • We need high storage USB keys.
  • We need to setup a credit card payment account.

Once those elements are in place, we’ll test the upload service and streamline it to “just grandma button” simplicity. At the same time, we can explore farming solutions. Once SAFE Network is up and running, we’ll add a SAFEcrypt APP to the process.

Farming Solutions

  • We’ll use the USB keys to plug in processing units (rasberrypi, laptops, towers, etc) to find the best configuration.
  • The goal is to run the most vaults per configuration, and discover the weakest links in the chain.
  • There’s always a chance someone will find a cheaper way to farm Safecoin, which is why we keep the option to refocus funds to purchasing SC directly off the exchange. Your ASIC miner purchase was the perfect example why we want to avoid going all-in on farming.

I look forwarding to co-op funding this business venture. After we establish good (farming + uploading) reputation, we’ll attract a large interest for public investment. People are more confident investing in something they know already works.

1 Like
#7

Here’s some stats using CrystalDiskMark to measure data transfer rates.

USB Flash Drive 2.0

Portable HDD USB 3.0

HDD 500GB Direct Connection

Obviously, the HDD drive is faster because it is directly connected to the motherboard. I suspect the USB port connection is the main reason for the slower transfer rate. BUT there’s significant difference between the 2.0 port and 3.0 port. The USB 3.0 is 5x times faster!

Preliminary Farming Theory

Let’s pretend a single vault gets 1MB chunk request every second. It’s an extreme magic number, giving us a place to start. Based on the stats above, we can estimate vaults per storage device…

  • USB 2.0 should be able to host 22 vaults.
  • USB 3.0 should be able to host 114 vaults.
  • HDD Direct Connect should be able to host 170 vaults.

This assumes the Operating System and RAM can process X vaults simultaneously. We’ll have to see how many vaults can run per operating system. Keep in mind, the vaults also have to do: caching, consensus, and routing.

I suspect the cap on vaults per Operating System (less than 100) will be lower than the vaults per Storage Device (more than 100). If that turns out to be true… it would be wise to use multiple OS devices with USB 3.0 storage plug-in.

You can only farm as fast as the slowest link in your chain.

This devices looks interesting.

http://www.buyraspberrypi.com.au/shop/4-port-usb-3-0-powered-usb-hub/

4 Likes
#8

Awesome! I love the stats. Would it be possible to also test USB 3.0 Flash? Thanks.

1 Like
#9

Yes, but I don’t have a USB 3.0 flash right now.

If you have one, just download CrystalDiskMark and run it while your USB 3.0 is plugged in. It’s as easy as selecting it from the drop down menu.

EDIT:

According to sites like Amazon, the USB 3.0 flash drive read speed ranges from 100Mpbs to 245Mbps. It depends on the flash drive…

1 Like
#10

Check out this blog on USB 3.0 flash drive stats!

This will help us determine which brands to avoid.

1 Like
#11

SAFEcrypt Farming

I have a friend who knows a lot about Networking hardware. I spoke to him about some configurations ideas. Surprisingly, he recommended using Western Digital (HDD) over USB 3.0 (Flash). The reason is because we are building a “mini server”. The read/write will be so numerous that it will wear out the Flash memory. It’s not designed for server functionality. SSD Drives are a different story, but the main objection was cost per storage space… that may change down the road.

I explained how SAFE vaults work, and he told me to find the resource requirements per vault (RAM, Processing, Upload Bandwith, Storage) and he’ll put together an affordable configuration to run 1000 vaults, on a 1Gbps pipeline, if possible.

SAFEcrypt Uploading

If our farming operation uses SATA 3 hubs, people can send their HDD & SSD drives and we plug it directly into our SATA ports. Quick and easy. This makes the process more efficient and requires the customer to provide the drive.

Obviously, we should be as flexible as possible, and also accept USB 2.0 & 3.0 flash drives. We should provide different options for them.

  1. Send us your HDD or SSD.
  2. Buy our SAFEcrypt USB 3.0 Stick.

Bandwith is Digital Oil on the SAFE Network!

MaidSafe is getting “vaults” ready for public testing.

What are vaults and why are they so important?

In a nutshell, vaults are the backbone of the SAFE Network. They do numerous functions to “secure” and “serve” Network activity.

Why are vaults important to me?

As a farmer (someone who runs a vault on the SAFE Network), you have the opportunity to earn Safecoin. In order to earn more Safecoin, you need to run more vaults. But each vault requires resources (processing, ram, storage space, and upload bandwith).

Why is upload bandwith important to vaults?

I believe upload bandwith is more scarce compared to the rest. Anyone can buy more hard drives, but it’s not that easy to get more upload bandwith. The more upload bandwith you have, the more vaults you’ll be able to support, which leads to more Safecoin rewards.

How do I get more upload bandwith?

Unfortunately, there aren’t many ISP’s, providing high upload bandwith. But there are ways to tap into this resource. Here’s some suggestions.

As a serious farmer, make the choice that works for you. If you’re hosting a farming co-op let everyone know. I prefer a local co-op because I can see the operation for myself.

What is a farming Co-Op?

Farmers share their resources (money, hardware, ISP access) to maximize Safecoin earnings. An example would be someone with access to Google Fiber, sharing the hardware and ISP costs with another farmer.

2 Likes
#12

There is one point that he may have missed

The vault that remains online is a write once read many situation. The chunk is written once and read many times.

For a decent USB3 flash stick, this is NO WHERE as bad as in write often/read often configuration for each chunk (of space)

I do agree that if you are going to be serious then WD HDD is by far the better option and the red drives are built for raid style heavy usage situations.

SSD drives for vaults will work fine and handle the load if a decent one. That is for the same reasons as I said for the flash, they are running in a write rarely read often application, rather than a file server application.

The IOPs required for 1000 vaults is what we will need to know to be able to configure and size the storage system/

tl;dr

Vaults run under a write rarely and read often profile, and flash/SSD run ideally under these profiles. (Assumes vaults rarely reset)

1 Like
#13

This was an oversight on my part, when explaining SAFE vaults. He just assumed it works similar to typical servers where data is constantly being deleted/rewritten. I’ll point that out the next time we speak.

Thanks for the feedback.

1 Like
#14

You’re welcome. I am a certified network engineer & sys admin amongst a number of others after all. But not actively working as one. So you should get better & more informed advise from your friend.

1 Like
#15

Cost Per Vault Speculation… MAGIC numbers!

Right now MaidSafe is using Digital Ocean to run a private test network made up of 50 vaults.

The cheapest droplet option is (20GB SSD @ $5.00 per month). Think of a droplet as a virtual computer connected to the internet.

I suspect 1 droplet = 1 vault
Obviously, I could ask MaidSafe for details, but where’s the fun in that?
If true, then MaidSafe vaults have a storage capacity of 20GB.

Digital Ocean
Total cost per vault = $5 per month

If we build a farming co-op for SAFEcrypt, we should keep pace with MaidSafe vaults. So let’s crunch some “magic” numbers. We will need to calculate 3 things.

  • Bandwith
  • Hardware
  • Electricity

Bandwith (Google Fiber or Similar Fiber Optic ISP)
A 1Gbps Google Fiber line costs $70 per month. So if 1 vault has a minimum requirement of 1Mbps, that means we can support up to 1000 vaults.

$70 / 1000 = $0.07 per vault per month

Hardware (WD HDD)
We’ll use a 1TB WD HDD drive as a starting point.
We can debate “Red NAS” vs “Black Performance” later, and also discuss 1TB vs 4TB. I’m using the black performance for this example.

If we can run 1000 vaults and each vault needs 20GB, that means we need 20 TB of storage space or 20 (1TB Drives). Each drive costs $70 x 20 = $1400.

Since the typical HDD has a lifespan of 3 years, we take the total cost divided by the lifespan, divided by the total number of vaults.

$1400 / 36 / 1000 Vaults = $0.04 per vault per month.

Electricity
This one varies from location to location. It’s really hard to guesstimate because you may also need “cooling” if your hardware runs hot. Right now, I have no idea how much electricity it costs to run 1000 vaults. But I’m going to try anyway, haha!

If the farming operation uses 1000 watts (my guess) and runs 24/7 for 30 days, then…
1000 watts x 24 hours x 30 days = 720,000 / 1000 = 720kWh per month

The average electricity cost here in NM is around $0.082 per kWh.

So a 1000 vault farming operation costs around 720 kWh x 0.082 = $59.04 / 1000 vaults = $0.06 per vault per month.

The actual electricity cost for the farming operation may be more or less. IF the total cost of your (home) vault per month is MORE than a (cloud) vault, you should consider cloud farming instead. However, I honestly doubt this will be the case for most people. Home farming with spare drives is more cost effective 99.99% of the time.

TL:DR

Based on the above “magic” numbers we have…

  • Bandwith = $0.07 per vault per month
  • Hardware = $0.04 per vault per month
  • Electricity = $0.06 per vault per month

Total Cost per Vault per Month = $0.17

I know what you’re thinking, that is INSANE compared to Digital Oceans $5.00 per month!
Yes, but it also depends if we can “reliably” run 1000 vaults.

Trust me, this is easier said than done. The main reason is dropped ISP connections. My home ISP drops at least once a month, which means my vaults reset and have to start over. If you’re a serious farmer, you want your vaults to stay “connected” for as long as possible.

4 Likes
#16

If my “magic” cost calculation are close to reality, then assume 1 SC farmed per Vault per Day…

This will blow your mind!

(1 SC) x (1000 Vaults) x (30 days) = 30,000 SC

If your monthly operating costs is around $170 (1000 vaults x $0.17) then you’ll be farming SC at $0.0056 per coin.

$170 / 30,000 = $0.0056

If MaidSafeCoin’s price is $0.05, you’ll be making 10 times!

2 Likes
#17

wow, this is fascinating stuff guys @MrAnderson … and the numbers look very favorable… I’m going to find some time to read up further :wink: If you’re looking at a co-fund/ co-farm idea then I would be very interested in principle … hoping for big news from David and team on Mon/Tuesday…

1 Like
#18

We’ll see how things go right? ethereum mining also has advantage you can mine ether for 50x cheaper than exchange price

1 Like
#19

Yes, i can see myself invest on farmig as well.

2 Likes
#20

yep, some way to go before we have any stable experience with Farming SC… but hopefully it’s nudging into sight soon ,)

#21

BUt also, I feel that adoption will be better if the price is much higher.

Rather than see some advantage like ‘zomg’ make tons of killings because of farming. That costs of farming are greater… rent of land, reliability etc. I think it rare you have 170,000$ to sink into farming and willing to do that. setting up 1000 machines is no joke, and keeping the up and running etc equally not a joke so there is that barrier to entry.

Also, the better the price the more people will be willing to farm.

What is the cost of all these human effort to make farmers active??

1 Like