For the most part Sub P-Rep nodes function as citizen nodes at this time. The ICON network is not under heavy usage at the moment, so a large number of citizen nodes provides unnecessary buffer to relay requests to P-Rep nodes.
Many teams are not technical, and forcing teams to run a node will take away their time/effort, which can be used to work on what they’re good at (e.g. marketing, bizdev, etc.)
The debate is between practicality and philosophy. Some believe that running a node is a core requirement to be a P-Rep (philosophical), some believe that running a node is a soft requirement that can be fulfilled if the team can do so without wasting time/effort (practical).
Many Sub P-Rep nodes are operated by OAKNODE. Depending on your views, this is a layer of centralization as without OAKNODE, these teams may not have the technical expertise or interest in securing a node.
Blockmove suggested a more practical solution like onboarding more active Sub P-Rep nodes rather than taking a governance-level action.
Unfortunately I was not able to participate in the weekly meeting so Im not sure what was discussed, but Im a bit intrigued by the suggestion for the node running to be a soft requirement.
I am just wondering - as there are currently sub-prep that have self-staked and whose sole contribution to the system is supposed to be running a node, what would happen if we remove that requirement? If I understand correctly, they will be getting prep rewards just for filling a form and that would be quite a waste of resourses imho
Wont a better solution be to work in the other direction - increase the amount of main preps, instead of lowering the amount of active nodes? Why should we put ourselves into future situations where we have to stomp fires again in the future by removing the requirements to run nodes now and then re-instate them when we need them once again and hope that everyone will react in time?
In regards to Oaknode - they are currently running a node for me as well. Im not technical enough to run a node by myself without encountering issues and it is currently a good option for me. The other option for me was to run a node with the help of Rob (as I think that also quite a lot of sub-preps did) and it would have been a lot cheaper like that, however at the moment Im not worried that I will have issues with the next update or if I have to upgrade to a main node at some point. I am also getting every node related request fulfilled as quickly as possible, the updates and fixes are tracked for me, they are offering multi-location set up, upgrade and downgrade whenever I wish etc etc - stuff that I would not be able to handle on my own…
I honestly dont actually see how this is different (in terms of centralization) from sub preps running a node with the help of Rob or using Rhizome’s script to set up a node. The only difference that I see is that if Rob or somebody else that helped me set it up is not online and willing to help at the exact moment that I encounter an issue with the node, Im pretty much screwed…
That’s a great point about the self-staked nodes. If there was no hard requirement to run a node, you’re right - self-staked nodes would essentially be a money printing machine and waste of inflation.
Can you clarify what you mean by…
Why should we put ourselves into future situations where we have to stomp fires again in the future by removing the requirements to run nodes now and then re-instate them when we need them once again and hope that everyone will react in time?
At the moment, there is no requirement to run a node as a P-Rep. Today’s discussion revolved around whether it would be a good idea to enforce active node operation for all P-Reps.
Regarding Oaknode, there is a clear difference between a managed operator versus having a friend (in this case, Rob) helping with setting up the node. This is similar to traditional web hosting where you have managed providers like ServerPilot, WP Engine, etc. versus having a friend help you set up a VPS on DigitalOcean or another provider. In the end, it’s about having root access and minimizing the potential parties that have access to your keystore file and password. At the moment, since Oaknode is mostly helping teams that do not participate in governance, I think that’s okay. If they control a non-trivial percentage of main P-Rep nodes in the long run, I’d be concerned as the nodes are likely spun up by the same template and would have the same vulnerabilities - that’s a topic for another day though.
Those are really valid points for Oaknode - I apologize if I did not understand the issue correctly.
If the key is exported and stored with the client (as this is something that is offered in the service, the only reson that most are not doing it is that ironically the key is less safe that way), would that still be a concern? Im actually asking as as I have mentioned earlier, unfortunately Im not that technical when it comes to blockchain and Im trying to understand the problem in full.
In regards to my ‘stomping fires’ comment - I meant that imho its not a great idea to formally agree that the node running is not required as at some point in the future we might get a sudden increase in transactions and might need the nodes to run. If we dont have that procedure in place by that point, this would mean that we have to react very quickly and hope that all sub preps will be able to react (also learn how to do so) fast enough so we can avoid major issues. Expecting something like that from preps that have not even bothered to spin a node as a contribution is … not perfect
There is also one more side to that - we do have enough sub prep nodes for the moment, but if node running is not a requirement, how motivated will the preps be to spin a node at all (and spend money and efforts on it) when they know that most others arent and the rest are still getting payed the same or even more for doing nothing?
Imo the more importatnt thing is that we have to stop with this ‘wild west’ attitude and implement clear rules and requirements for all preps (some of them could be separate for sub/main preps) and we make sure that there is a strong guideline on what is not only expected but also required by the preps as at the moment everyone does what they feel like and the only effect by this is that some preps are using that as an excuse to do nothing, while others are carrying the full weight of supporting the chain for no extra benefits (and often more expenses)