after watching the Q2 Town Hall video I have some questions about the “Commercial node operators” and “Data center partners” because it seems not to be in line with the decentralized storage concept of Storj, “do not buy dedicated hardware”, blabla… but it is really interesting.
Are the SLAs of a “commercial node operator” already defined?
Tardigrade satellites will have any preference for these commercial node operators?
Commercial node operators = data center partners? if so, is there any limit in storage capacity of the datacenter?
I took it as they would provide tooling to more easily manage multiple storage nodes. Not as something that would require SLA’s or preferential treatment. I’m assuming these nodes will be treated no differently by the satellites. So you would still have to make sure you’re running on different /24 subnets. But lets say someone has a large infra to their disposal and wants to spin up nodes on all separate /24 subnets in different locations around the world. Manually updating, monitoring and fixing those nodes could be a pain. If there is a more central way to do that it could help with commercializing such distributed setups.
I could of course be wrong, but I feel it would be a really bad idea to create different classes of storagenodes.
I hope people don’t watch the video and think I better go rent some servers from datacenters around the world now, Thinking there going to get paid more.
Agreed. I love the concept of storj and how easy it is to run, but if it starts granting data to a commercial SNO over a regular SNO then I’m out. No sense in supporting a “half-centralized” storage platform prioritizing datacenters.
My Node only gets half traffic caused by another Node somewhere out of my ISP Subnet…
So every month i`m getting punished for that. If Datacenters does not in that case i think the Node churn could be huge.
Nodes at data centers probably are going to get more effective ingress traffic, but not because of any preferential treatment by Storj. They’ll have bigger pipes and faster storage devices, so they’ll win a larger share of upload races. (In fact, they already do.)
This is a problem. On the flip side of this coin is that most data centers have many IPs at their disposal, so if that’s the main way we are trying to prevent it, i don’t think it’s that effective.
The one of the goals - to do not store pieces of the same segment on nodes with the same ISP, include datacenters.
It’s effectively distribute pieces to different locations.
Datacenters are an inevitability no matter how you look at it. Look at what Bitcoin mining (as well as all the others) has turned into. There’s huge “datacenters” (or more appropriately mining farms) all over (primarily China) mining everything now. It starts out with the little guy, and the only thing that truly keeps most people interested is the profit… and anything that’s profitable as well as “farmable” WILL be farmed… period. On the bright side though, it will be “centralized decentralization” in the sense that no datacenter will have all the data in the way typical datacenters do now. The Storj network will still have all the benefits as well as even better network throughput and reliability, and yet anybody can still build their own datacenter limited only by what their willing to invest. Also, the long term success of Storj kinda depends on this inevitability as there’s no other way to be able to handle the exponentially growing demand for data storage. Your average person hosting a hobby node is not going to be the one supporting this or similar networks in the long run, but that doesn’t mean there not still imporortant or profitable. As long as the numbers remain consistent, small nodes should be just as profitable as datacenters relatively speaking based on their available capacity and bandwidth, so I don’t expect datacenters would hurt anyone similar to how mining works. Digital currency mining is literally just a race based on probability related to your limited piece of the pie, whereas Storj is… do you have some space and bandwidth to handle some of this exponentially growing pie? Right now it’s more of a waiting game for people to get onboard to decentralized storage. Once that shift really starts to happen though, I truly believe most won’t be able to expand fast enough.
…