I have more examples, how to make it more expensive 
You may run your main load in one provider (Azure for example), and run gateway-st on another one (AWS or GCP).
You will pay for egress to Azure when you upload to the network, pay to AWS (because it’s now egress for them) and pay for downloads: to Storj for egress, to AWS for egress (to Azure).
1 Like
UFO arrived and left it here: for fun and profit
1 Like
I think now it’s the time for RAID owners to rethink the setups. I can’t see worth running more than 1 HDD per node.
Also, who wants to stay, should consider moving the data from small to bigger HDD, like 20TB. There will be no gains from under 6TB drives.
5 Likes
Lol. All the more reason to offer free bandwidth. But it should be done on a per customer basis depending on customer requirements. This way as Storj grows and the technology matures these free packages can potentially be scaled back. Customers would be more than willing to pay more than $4/TB as it would still be far less than competitors as long as they can save on the bandwidth side of things. But even at $4, they’re not goinng to switch if bandwidth costs will increase.
For honestly, I think it’s better to convert a free tier to a REAL trial tier. Like - “you have 150GB storage and 150GB/mo bandwidth for 3 months, then you need either become a paid customer, or go away”.
17 Likes
Stop with this free engress. If it wasn’t for the engress earnings, I wouldn’t even started storagenodes. The engress is the main money maker for SNOs. Who dosen’t aprove that, he dosen’t run nodes. And don’t tell me you have some 16-18 TB nodes that are already filled up and you are happy with them, and don’t look to expand.
Good point @Alexey. Free accounts must be temporary.
1 Like
I actually just made the switch off raid pools partially for that reason. Otherwise I would have had to spend more $ optimizing the pools.
That’s a good idea for the free tier, but it doesn’t do anything to attract the bigger players. They couldn’t care less about the free tier. The free tier is good for devs or people that want to simply test out the network without having to pay for that testing, or even everyday people just looking for a cheaper backup option. It does nothing to offer incentives to attract big customers.
Why do you believe customers always have a usecase that would not be perfectly even around the clock? Btw my grafana dashboard doesn’t show perfectly even traffic.
Beside that no we don’t manipulate the numbers. We splitted customer and test satellites for a reason. This has not changed.
3 Likes
I would downgrade my internet connection and try to get as many used space as possible without worring about the download performance for the customer. I will have to optimize my setup for cold storage at that point. I don’t think the current customers will appreciate that.
Same for the idea that download traffic should be free. I am fine with that. I will adopt to the new incentive. Customers will just have to wait longer for the download to finish.
How would be $0,75/$1,5 (storage/egress) better?
My nodes average about $4/TB accounting for egress, however I have plenty more bandwidth that’s not being utilized. It neither makes me more money or cost me any more to have. At least for the time being, we’re facing cuts to some degree. The proposed cuts will at minimun bring those earnings down to under $2/TB. So… I don’t think a flat $2.50 to $3/TB stored while also giving Storj a huge selling point to attract new customers is an unfair compromise. SNO’s would also have a steady predictible income and all the crying about Storj increasing egress would be moot!
Now, I say per customer packages to obviously limit this being abused and allow for things to change in the future as other outside factors change. Any overages would obviously be charged at a fair rate and should be paid to SNO’s as well. Please explain to me how that’s in any way worse than what’s been proposed by Storj here?
2 Likes
to attract big players we actually have what to offer:
- global distribution;
- great SLA;
- simple migration (s3, you know, but I personally think it’s fair to charge more for s3 without running your own S3 gateway);
- end-to-end encryption;
- plain prices (oh, maybe segment fee
); - fastest throughput (especially in a native integration and some tunings);
- lower prices (this one is the last what I would account if I’m a manager who able to make decisions).
4 Likes
its certainly a good point… if egress doesn’t pay one could simply have near infinite data and no real egress speed… which would be bad for customers…
but egress earnings also go quite low when the data gets old…
so not that huge an incentive for relying on egress
already…
ofc we all want to earn more, so its good that an egress incentive exists and a good argument for that it should keep existing for sure.
In many cases this probably doesnt matter much since most connections people have are already the cheapest options they have and Storj still works fine. If it did become an issue though, easy. Storj can impliment a speedtest into the nodes. The more bandwidth you have available the more data your allowed to store. Take it a step further and “cold” data can be stored slightly more on slower connections while “hot” data on faster ones. Sure it would be a little work for Storj to impliment this but isn’t that kind of the point? To continue to do things to better optimize your product and create incentivizes for customers to use it?
Trust me when I say… currently the ONLY people who MIGHT be paying for a higher tier internet plan for the sake of Storj nodes are the whales. NONE of your average users are paying more for their internet specifically for the sake of running a Storj node!
And if we simply waive repair costs, Storj could more easily shuffle data around as needed to best optimize this. That way people with slower connections aren’t at a major disadvantage. Of course the network could know ahead of time using speedtests what nodes have what speeds available and distribute data accordingly in the cases where Storj knows ahead of time what customers will use a lot of bandwidth based on their bandwidth packages. Everybody wins!
If the above is implimented, this probably wouldn’t be necessary.
1 Like
Or the next lower bandwidth option is so bad that you simply would not do it. This would be true for me here.
1 Like
That doesn’t work. The satellite would have to perform these speed tests. I am able to modify any speed test that would be executed client side. Even for a satellite side speed test I would modify my node in a way that the speed test always has fastest possible speed while slowing down all customer downloads. You can try but I am confident that will not be a big problem for my node.
If download traffic is free why would anyone want to store hot data? I expect all nodes in the network to optimize for cold storage.
Call me a whale than! I am already looking into downgrading my internet conection 
2 Likes
Do you really think that this is gonna make a real difference? I mean every free account is getting a discount of max $ 1.65 per month. Care to share how many free accounts are there?
1 Like
Paid accounts also get the same discount. → You can take the total number of accounts from the public endpoint.
1 Like