So, I’m employed by a multinational hospitality company with 40,000 employees and currently engaged in a £60 million, 5-year IT services contract with Fujitsu, (just to provide context on our scale).
I’m curious about Maidsafe/Autonomi’s perspective. Do they anticipate businesses like this opting for private networks (will this be possible?), or is there a different focus? It seems to me that while companies might find value in a private network, the real benefit lies in them embracing the actual Autonomi network itself. After all, the widespread adoption by companies would benefit the network more than the creation of isolated private networks and as you imply @TylerAbeoJordan also be much simpler for the company.
So, could / should a business theoretically establish its own private or mini autonomi network encompassing all the laptops, computers, and mobile phones in use in the business?
And if so, would this network need any linkage to the broader network for upgrades etc? or could it operate entirely as a self-contained, private entity? Would there be any benefit to this given how private and secure autonomi will be…?
It shouldn’t differ in the same way if you were to simply clone the Bitcoin code and use it strictly in-house.
You would have control over the entire supply, and all private keys (essentially)
So were you to clone the entire Autonomi code and simply run it in all your global locations as a different form of VPN + SAN… You would have full control over the tokens generated and used.
You could choose to have it interoperable with the original Autonomi network, but you would have to change code so that your tokens are discernible.
Or…at least I imagine… you should be able to have it be fully funtional as a private internal network.
I guess the benefit might be, replacing expensive VPN servers/licenses AND storage server operating systems licenses
Edit: It should be noted, you’re not realistically thinking about doing any of this with actual business systems for at least 5 years after the network has been live and has proven its security and resilience to attack and other malicious entities/activity.
Appreciate your input. Your analogy between setting up a private Safe network and using existing blockchain code for internal operations is good.
I’m curious about Maidsafe/Autonomi’s perspective. Do they anticipate businesses opting for private networks (will this be possible?), or is there a different focus? It seems to me that while companies might find value in a private network, the real benefit lies in them embracing the actual Autonomi network itself. After all, the widespread adoption by companies would benefit the network more than the creation of isolated private networks and as you imply @TylerAbeoJordan also be much simpler for the company.
You need to remember the network is being designed to solve untrustworthyness at very large scale.
In a business, it’s pretty much assume there is trust on the inside, so already a vast majority of the codebase is unnecessary.
Also what Tyler mentioned about having a large amount of nodes, that is also what the network is designed around, having a huge amount of disparate nodes to stabilize and make the storage redundant.
I’m sure David will go into crazy detail with these things, I can’t get close.
A lot of businesses are going to zero trust architecture. A big security issue is if the “shell” gets punctured with the “inside is trusted” architecture, an attacked has free reign inside the network. With the zero trust (inside is hostile too) it is much more robust.
All of that to say two things I guess:
don’t get rid of the security workings of the network because you’re running it “trusted”. It’s much less secure.
the way Autonomi is designed, you should be able to make your own little enclave “in the open” / on the public network and have it be just as secure, if not more secure because it is handled by a larger network, than trying to do it yourself on the smaller scale.
So from an IT services standpoint, it seems a partnership with a company like Fujitsu, as seen in my example, might be more beneficial for Autonomi than working directly with the hospitality firm. Because established IT service providers like Fujitsu have the resources and expertise to integrate innovative technologies like Autonomi effectively…
I think if you set-up multisig encryption layers for different groups, then just use the main network. Not sure entirely how this would work, but main net will allow multisig at the base layer, so with private data any group you set up can already use and share info that way. Adding another private encryption layer onto that somehow might get you what you need where you can have groups within groups. I don’t know much about multisig though and adding/removing members from groups may be a problem?
I do remember several years ago maidsafe talking about one of its business possibility would likely be assisting execution and support of private networks. That should always be a possible for the company that created and launched a successful antonymous network as Autonomi should be. But that would come much later, after so much more is done.
In the meantime, there’s nothing stopping others from using the code. Understanding it fully is another matter altogether.
I mean it could be done, and work but its not what its designed for.
I cant see updates not running.
If it were to be run closed you lose the whole need for payments.
It would need to be run at a very large scale to offer the same security as the main network, so if thats needed why not just join the main network and do it at a cheaper expense.
Imagine 1/3 of a private network going down to a ISP outage. Due to its smaller size you’d have a huge churning event. Would the remaining 2/3rds be able to handle the duplication of chunks and storage requirements?
I think it’d be better to go build on top of Autonomi.
While 60 Million pounds sounds like a lot of money, most of that will go to network infrastructure, software licenses, security, software licenses, helpdesks, software licenses, technical delivery staff, software licenses, hardware, software licenses, consultants, software licenses… did I mention software licenses ? Smear it out over 5 years, and 60M doesn’t look that astronomical anymore.
Autonomi could only play a very limited role here, maybe as an object store or a backup dumping ground. You can’t run a database or a virtual machine off autonomi, to be fair they would be better off with some shared san storage in a fujitsu managed datacenter somewhere or maybe a software defined storage solution if they actually have lots of data.
Businesses like that usually have a couple hundred TB in storage needs.
The enterprise IT world will be stuck with cisco, vmware, redhat, citrix, microsoft, sap, oracle and netapp for at least decades. Or they go to the public cloud, and will be kept hostage there forever.
For the immediate future, you are certainly correct.
On a not-to-long timeline, as the autonomi network grows, things should change pretty rapidly, I think. Virtual machines will need compute resources which is a way ahead, though envisioned.
The key things that will be supplied rather immediately are (hm, let me see):
–secure, permanent storage and retrieval of private data (unsnoopable) at a really, really low price.
–uncensorable, permanent public data
–communication which are as anonymous, pseudonymous or identity-proven as the user chooses.
–wayback machine always on, for the life of the network (and likely beyond: With archive nodes functioning and the ability to re-bootstrap the network, even if dead, the network could be reanimated.)
–completely secure p2p connections set up privately for full direct computer to computer interactions. (might serve the MMOG community sooner than later?)
There’s certainly a lot more, but with some of the basic function set up I guarantee the further hurdles to expanded function will fall quickly.
What we’re looking at here is truly game-changing–a different landscape emerging. Recognition of that will come as it comes. In the meantime and always, players will do what they think best serves their interests.
Yes, this is not only possible but relatively easy to do. The hard part will be the deployments of nodes, what SOE will be decided upon to specify which machines runs nodes.
The network has its own key and nodes will only talk to those using the same network name/key, so not only do you have the protection set up in the corporate network but also the nodes only talk to other nodes in your “private” autonomi network.
I don’t think that autonomi will ever be a drop in replacement to just keep working the (dare I say rather stupid) way we work:
Bob opens up his laptop. Starts his vpn client and citrix client which opens up a remote connection to a virtual desktop that is running on a hyper-v virtual windows server somewhere in a datacenter far far away. On his virtual desktop, he opens up his browser and points it to the web frontend to SAP that Alice coded (years before she left the company). That web frontend is running on a virtual server, and it connects to the SAP virtual server, which in turn transacts with the virtual oracle database servers (which does not run on hyper-v but on oracle VM because of licensing cost).
To build this business logic on autonomi needs a whole new way of computing. The computing needs to be brought back to the computer/browser, but with autonomi as a data backend.
Honestly I doubt autonomi will replace high speed database engines. To get millisecond access to data will not occur on automoni which is needed for some applications. Cern is not going to dump raw data to autonomi, not when they are filling many (a lot) of diskdrives in a single run of the hadron.
Autonomi will certainly though fill the market space beyond those applications that demand high speed access.
True, CERN has massive ceph clusters for that. As far as I know ceph is the only storage solution that can do storage at that scale and at those speeds.
I don’t think autonomi will ever beat ceph at that game. Right tool for the job and all that.
in-memory databases provide this millisec access to data service already, checkout Hazelcast, they do tap into backend disk based data stores structured and unstructured. Redis, Oracle Times Ten do the same (cisco uses this inside their routers), there are many already employed to deliver such high speed service, In theory Autonomi could be a backend disk based data store for any of them…
as the storage backend.