Do you plan to have an easy migration path from one model to another, let say, moving data from a global/regional workflow to the archive model? Like, maybe through S3 lifecycle rules?
1 Like
This is a big if. Why would vpn slow node? It adds one hop on the very low latency network and depending on provider may actually result in better routing and overal better performance.
I don’t buy this whole “people cheat with vpn” spiel. For one — why would anyone sell their integrity for literally a couple of extra bucks a month, and another — it’s trivial to detect by correlating traffic latencies, let alone downtimes. Sounds like a pretty dumb idea.
The miniscule amount “cheaters” can be ignored. They then have $5 and no self respect. No harm to the network. Remember, nodes are already assumed Byzantine.
1 Like
With this pattern perhaps a better tier would be an Archive, it doesn’t have a segments fee, and looks like your bill wouldn’t change.
Right now I think it’s tied to a bucket or a project level, the migration may be done with a repair, if there are different expansion factors. If they are the same, it’s just a tag for billing.
Not sure that the migration between global and regional can be always performed with a repair. In some cases you would need to re-upload the data.
1 Like
Hopefully I’m wrong here, but migration from legacy to any new tier seems like the bucket/project will need to fully transfer to a new set of nodes or something like that. On the date that all legacy tier users need to transfer, how would that be handled?
Yes, we are looking at supporting movement between tiers either on a per-object basis, lifecycle rules, or something else.
As Alexey explained we do already have some limited features in repair to handle similar things but there is more work needed to handle tier movement.
2 Likes
We are hoping to not need to do so. I can illustrate how based on capability we have today. For example, if moving from Legacy Global → Global collaboration is a large increase in number of nodes storing the data (increase in expansion factor), our existing systems already can handle that bump while keeping all the existing pieces. We are finalizing more details about Global collaboration so I hope to be able to share more in the coming weeks. Because it is not finalized yet there is a still a chance we land on a plan that might not keep the original pieces. Again, we are hoping to not need to do so.
2 Likes