MasterNode Scaling...

camosoul

camosoul

Well-known member
MNOs, are you going to build you own datacenter?

When the chain scales to 100s of MB per block, and you're required to store it on an SSD (array), are you willing to do this? Are you going to build you own datacenter and fill it with this junk?

I'm not a trader/ponzi/speculator, but this forces one to ask; do you think DASH's price will ever be high enough to support what it demands?

Doesn't this result in centralization?

Will each MN have to have it's own, unique copy of the data stored? If not, does this consolidation harm integrity/resiliancy; also play to centralization?

Will DashDrive shard data?

If DASH does get that valuable, won't we then face a granularity/resolution problem? Perpetual huge blockchain on Masternodes seems to fix a scale-ability problem, but it would need to be of such a large value that it creates a fungibility/resolution problem. Are private sidechains inevitable? Will they be accommodated? Doesn't IX/IS pretty much accommodate them?

WORM/ROM/PROM/EPROM/EEPROM version of SDD? Cheaper or more expensive? SSD-flush-to-WORM when it doesn't wag for X blocks? I don't know the FS that does that... Is this the special hardware Evan want's to make?

People with brains; what say you?

Surely, dirty hippies will want to blow up your building. Can you, will you, defend it? What if those dirty hippies are neo-Bolshevik guvthugs pushing StateCoin?

DASH inevitably becomes non-portable. Will you build in Florida? Or Utah? Or the Sahara Desert and power if from PV? Won't this result in geographical consolidation? Everyone will pick the same geographical/political stable areas.

Is it even possible to remain distributed and trustless? Should we just give up and accept the inevitable?

Going forward, MNOs need to know... The MNOs of today are largely too stupid to ask, much less answer these questions. But, what few brains there are, what say you?
 
Last edited:
We're still very far off from that.
But personally, I'm preparing to have 1Gbit net and a fast computer for MNO, going into 2018.

Don't forget that hardware is progressing very fast. In a few years, we will have massive SSD drives that are not that expensive.
And yes, I would build my datacenter if the network requires it and I can afford it from the MNO rewards.
 
I think the days of universally distributed blockchains are gone and central blockchains will take over. I believe miners are dead and masternodes will reign supreme, the fewer the better. I believe instant transactions will be the only kind of transactions, why would you settle for less?. I believe total anonymity and privacy will win in the end. I also believe fees would go the way of the dodo. And lastly, I also believe there will be two main currencies (and a thousand wannabes), one for store of value and one for global commerce. That's the path of less resistance, and resistance is indeed futile.

It won't happen overnight but that's the future as I see it. Adjust accordingly.
 
I don't see this as black and white. Some cryptos will be far more distributed but unable to meet the demands of high tps. Some cryptos will find their niche and not target mass market. Dash, therefore, is filling a gap that would of been filled anyway (mass market with ability to store high value). Technologies like DAG may work for low value transactions but at this point I am skeptical they are secure for high value.

I understand where you are coming from. The only thing I would say is, and I suspect you might agree, is that dash needs rapid expansion of it's R&D to face these challenges. We either innovate and push the boundaries, or die. Tbh, I could make that argument for all cryptos.
 
There could be minimum requirements on MNs that the software can enforce. If MNs dont meet those requirements then no pay.

BTW I think Dash should look into sharding too.
 
camosoul said:
MNOs, are you going to build you own datacenter?

When the chain scales to 100s of MB per block, and you're required to store it on an SSD (array), are you willing to do this? Are you going to build you own datacenter and fill it with this junk?

I'm not a trader/ponzi/speculator, but this forces one to ask; do you think DASH's price will ever be high enough to support what it demands?

Doesn't this result in centralization?

Will each MN have to have it's own, unique copy of the data stored? If not, does this consolidation harm integrity/resiliancy; also play to centralization?

Will DashDrive shard data?

If DASH does get that valuable, won't we then face a granularity/resolution problem? Perpetual huge blockchain on Masternodes seems to fix a scale-ability problem, but it would need to be of such a large value that it creates a fungibility/resolution problem. Are private sidechains inevitable? Will they be accommodated? Doesn't IX/IS pretty much accommodate them?

WORM/ROM/PROM/EPROM/EEPROM version of SDD? Cheaper or more expensive? SSD-flush-to-WORM when it doesn't wag for X blocks? I don't know the FS that does that... Is this the special hardware Evan want's to make?

People with brains; what say you?

Surely, dirty hippies will want to blow up your building. Can you, will you, defend it? What if those dirty hippies are neo-Bolshevik guvthugs pushing StateCoin?

DASH inevitably becomes non-portable. Will you build in Florida? Or Utah? Or the Sahara Desert and power if from PV? Won't this result in geographical consolidation? Everyone will pick the same geographical/political stable areas.

Is it even possible to remain distributed and trustless? Should we just give up and accept the inevitable?

Going forward, MNOs need to know... The MNOs of today are largely too stupid to ask, much less answer these questions. But, what few brains there are, what say you?
Click to expand...

It is not the problem because if Dash block up to 100 Mb then its price should be about $10,000. If I cannot do, I will find and hire someone who have enough technical competence to do that for me.
 
It is good to be thinking way ahead and I think @camosoul makes a good point, but for me the answer is yes, I would be willing to do those things. These changes are not going to happen overnight so it's not like it is going to be a sudden shock to everyone. Over time, people who are overwhelmed or unwilling to do what is needed for MN upkeep will eventually stop operating MNs. Whether they also sell their collateral as well is a separate decision.
 
Well thier is a high level colo right down the street from me I can get a rack with 30 amps 100 MB’s synchronous internet connection over 6 backbone providers for about 7500 a month so dash goes over 1k I can pay the rack fee every month. So we get dash to 5k sure it would be easy to afford the requirements.
 
There is a possible solution to scaling that doesn't involve quadratic increases in difficulty with linear increases in transactions: hashgraph. Evan Duffield's announcement of the scaling roadmap, if implemented, will require masternodes to make significant upgrades in hardware and bandwidth. This may work for a while, but because verification difficulty increases quadratically with linear increases in rate of transactions, eventually will produce diminishing returns and possibly masternode falloff. Licensing hashgraph technology will could potentially increase transaction capacity dramatically for less money. The tech has a U.S. patent, so a 20 year lisence is effectively the same as a permanent license. Even if we didn't end up adopting Swirld's nuclear blockchain on steroids, it would still be worth the investment ( prolly a few million $) as an insurance policy.
 
Ftoole said:
Well thier is a high level colo right down the street from me I can get a rack with 30 amps 100 MB’s synchronous internet connection over 6 backbone providers for about 7500 a month so dash goes over 1k I can pay the rack fee every month. So we get dash to 5k sure it would be easy to afford the requirements.
Click to expand...

Wow that's really overpriced. Check out Hurricane Electric.


Sent from my iPhone using Tapatalk Pro
 
Is thier internet configured to failover to a different provider automatically and ur static ips stay the same?
Do they guarantee your bandwidth?
Do they have power from 2 different transfer stations?
Are they a tier 3 certified data Center?
You want a high level dc if your gonna be running a mn once all the requirements require you to have a rack my experience is you pay for what you get with datacenters.
 
Ftoole said:
Is thier internet configured to failover to a different provider automatically and ur static ips stay the same?
Do they guarantee your bandwidth?
Do they have power from 2 different transfer stations?
Are they a tier 3 certified data Center?
You want a high level dc if your gonna be running a mn once all the requirements require you to have a rack my experience is you pay for what you get with datacenters.
Click to expand...

From their wiki page,

Within its global network, Hurricane Electric is connected to more than 156 major exchange points[8][9] and exchanges IP traffic directly with more than 6,750 different networks.[10]

According to European Internet Exchange Association (Euro-IX), Hurricane Electric ranks first in the world for the number of connections to Internet exchange points, with presence at 80 of Euro-IX member IXPs.[11]


Sent from my iPhone using Tapatalk Pro
 
Better hardware, better software, and better decentralization. There are ways to incentivize better decentralization/distribution of the hardware as well.

Yes, I accept your challenge.
 
Koromoto said:
I think the days of universally distributed blockchains are gone and central blockchains will take over. I believe miners are dead and masternodes will reign supreme, the fewer the better. I believe instant transactions will be the only kind of transactions, why would you settle for less?. I believe total anonymity and privacy will win in the end. I also believe fees would go the way of the dodo. And lastly, I also believe there will be two main currencies (and a thousand wannabes), one for store of value and one for global commerce. That's the path of less resistance, and resistance is indeed futile.

It won't happen overnight but that's the future as I see it. Adjust accordingly.
Click to expand...

Stop taking the pills. Start knowing ...

hqdefault.jpg
 
Last edited:
I've been spending the last few weeks really combing through the information (Whitepapers, Videos, Articles, etc) about Hashgraph, and given the huge amount of funds available to us through the Treasury at this point, I think we need to seriously consider integrating the Hashgraph on the Dash Network. It would immediately allow us to scale to greater than credit-card TPS, greatly reduce the amount of data we would have to send, and drastically increase total network security. Obviously this would be a huge overhaul for the whole system, but we'd be the first to implement the Hashgraph on a public ledger system. Honestly, I think it's a brilliant move that would effectively future-proof the Dash network.

Even at this level, it completely blows all blockchain based systems out of the water and it's mind-boggling that no one has managed to implement it in a Cryptocurrency yet. We have the ability to become the first, and I think we should.
 
I don't think that speed will be an issue for Dash for a few years at least, also to implement such a big change would take a very long time and would require extensive rework and testing, but I agree that we should consider all of the alternatives for the future.
I would like to hear from the core team if there is someone that is looking into this and similar ideas?
 
You must log in or register to reply here.
Back
Top