Hi!
How can I speed up the “Graceful Exit” process on nodes? I have a 10 Gbps upload channel. I want to finish the “Graceful Exit” as quickly as possible.
I am not sure this is possible as:
1 Like
Graceful exit will take one month (if successful,) there is no way to speed it up. After one month, if the node´s online score is at or above the current online threshold set by the satellite and the node is not suspended, contained, or disqualified, the node exit will be considered successful, and the held amount for the node will be released with the regular node payment process. Otherwise, the graceful exit will be considered failed, and the held amount will not be released
1 Like
If you need space now — stop and delete enough nodes to recoup enough space. Waiting 1 month to get bac $10 (on which storj didn’t even pay interest) is a terrible financial decision
Maybe I will. You are right. It is very humiliating to wait a month for 10$. STORJ is not worth any attention at all.
Maybe this period could be reduced?
Waiting 1 month with the current GE impruvements is somehow abusive.
The pieces are not moved from the said node, but repaired from the network, so the process is way more quicker. Way wait 30 days? Usualy GE is done because you want to get rid of the node quicker, maybe hardware is failing, maybe you move out from your home, etc. Please revise this process.
It cannot be reduced at the moment. This period is needed to make sure that all pieces are repaired in time.
The repair queue is never empty.
It could be possible to reduce (or increase in some cases - depends on the stored amount), if we turn back to the direct transfers p2p from the GE node to several others. Unfortunately, it doesn’t use DHT or similar discovery in traditional p2p, the list of nodes should be prepared by the satellite. This is a resource consuming process, also direct transfers were often a result in disqualification, because the source node cannot transfer pieces with configured settings or problems with the router (it becomes overloaded), or the target nodes are unable to receive a piece or just failed in the middle of the process and so on and so forth.
Checking all segments to figure out all pieces on GE node and check their health even more resource consuming process, so this is again a probabilistic model as with audits, otherwise it will be very costly.
4 Likes
According to stats from stats.storjshare.io less than 10 percent of all exits are graceful. So I don’t think GE is that important.
1 Like
It’s probabilistic and works like insurance. If 90% of nodes would graceful exit perhaps somewhere something can be tweaked to increase profitability of the network.
But yes, this is an indisputable confirmation that holding SNO money hostage as an incentive (lol) for staying in network is not working. But since it’s been like this forever, either every other alternative is worse or nobody cares to evaluate this behind other priorities.
I’d argue it’s not needed at all. If people stayed for a year to make any sort of reasonable money they will likely stay longer anyway. And again, if they need space back — $10/3 years will hardly give them pause.
Maybe it’s there because it allows to tell prospective customers “SNO have financial incentive to stay in network” because this customers will understand. To grasp the mechanics and math behind not actually needing it and how network of crappy and flaky nodes can actually be ultra performant and reliable requires some specialized knowledge
1 Like