As one of the Altair collators who has been onboarded in a first batch, I would like to start discussion about right size of the collator rewards and its influence on overall quality of the network.
The current monthly reward is 2,500 AIR which, translated into USD with AIR 90-day average price, is around $25 per month. As you probably know, this can’t even cover the basic infrastructure costs for node operators, let alone having a profit for the service to the network. The consequence is that a longer this bear market takes, the more will infrastructure operators will be forced to cut the costs of their operation if they are not profitable. This can subsequently lead to the degradation of the network performance in a form of missed blocks and other issues since collator nodes can be run on a sub-optimal hardware equalize for a low reward.
So as for the actual numbers for adequate rewards, I propose that each collator should write down their actual operation costs to kick off the discussion and we start it from there Also there’s a thread on Polkassembly started with me and Paradoxx discussing the pricing of validators and collators so it might be useful here as well.
Anyone is certainly welcome to join the discussion, feel free to leave any comments!
Hello Pmensik!
Glad that you initiated a separate topic regarding collators’ rewards discussed so far in the Block rewards thread. I think this is the first and very important governance step.
Does this cost strictly related to running the Altair node?
I-m not definitely deciding the number of rewards that each of the collators will earn, but I`m curious to know what you would like to propose. Any ideas? Numbers?Proposals?
Does this cost strictly related to running the Altair node?
In this case, yes, we chose appropriate hardware based on the activity on the parachain - obviously Moonbeam will have different requirements than Altair in its current state.
bare metal - $50 (need to move providers soon though)
service and maintenance $50
going with the 25$ flat for monitoring and reporting
backup is another 50$
that’s 175$ costs per month while getting 2500 Altair
in my opinion we should get at least operational costs covered by the rewards for running the collator
maybe it could be done in a way where the rewards could be adjusted based on the price of Altair on the day of the payout. Since the collator rewards are coming from the treasury this shouldn’t be a big problem
Thanks for this Request for Comment and the transparency on the requirements and costs for running a collator node.
Being a collator on Altair is one of the pillars for stability in the ecosystem and everyone who does that should be fairly compensated. Based on these terms I fully second your proposal, I just want to mention we need to find
(1) an amount which suits all other collators as well and (taking the price volatility into account) we
(2) need to make sure the amount is sufficient for the next rounds too
Thanks for creating this topic @pmensik. Agreed that the 2500 AIR is nowhere near covering the costs for the infrastructure. But we have
We are operating on AMD Ryzen 5 3600, It uses avg. 6% of cpu, less than 1% of I/O nvme and 5% of the memory. It’s a good machine for parachains where also little performance is required. Also in the future we will use relay-chain from rpc where the parachain can even have more capacity to run.
So, without a backup, that accounts for $155/month so let’s round it to $200 to count in some actual profit for running the service.
I think our total is around $30-50 lower. Given by the example above that would be around 20000 AIR per month at this time of writing. It would cover the costs but it could create a lot of unwanted effects in the Altair ecosystem. Altair so far just has a NFT store, most of the functionality still has to come through, but also it would be nice to hear from the development team to see how the efforts are divided among altair-centrifuge.
I personally think that actual rewards can not cover the tech cost and Altair Collators rewards should be increased, but as you mentioned Altair Rewards can not be increased over 150-200$ a month due to the impact on Altair Ecosystem.
Once Centrifuge Treasury will be funded you can start the discussion or start directly with the proposal to run RPC.
Great point also about the tech specs of PC and using.
I`m not a tech expert but running an Altair node with Dual Intel Xeon Gold 6226R - 32c/64t - 2,9 GHz/3,9 DDR4 192GB/768GB does not make any sense…
I personally think that actual rewards can not cover the tech cost and Altair Collators rewards should be increased, but as you mentioned Altair Rewards can not be increased over 150-200$ a month due to the impact on Altair Ecosystem.
What do you mean by the tech costs? And for the 2nd part, are you saying that rewards cannot be higher than around $200? I am not saying we should get more, just wonder what’s the reasoning
Also, IMHO it does make sense to run nodes on a higher spec machine until the team says otherwise - we have to protect our reputation as a collators and being accused of running nodes on some shitty hardware wouldn’t definitely help that
Ahoj Yes, that’s why I created this thread, to see what the operational costs of other collators are so we can base reward adjustment on kind of an “average”. Also, I believe we can re-evaluate let’s say every 90 days so the rewards are not too low or too high
We are talking about Altair. Just now, as you already mentioned the Altair collator’s monthly reward is 25$.
Yes, I think that 25000 AIR as a monthly reward for Collators is not acceptable right now.
25000 AIR * 10 collators = 250 000 AIR /month
Or 3000000 AIR a year.
I highly doubt that this reward rate will benefit Altair Ecosystem.
We are talking about Altair. Just now, as you already mentioned the Altair collator’s monthly reward is 25$.
Yes, I think that 25000 AIR as a monthly reward for Collators is not acceptable right now.
25000 AIR * 10 collators = 250 000 AIR /month
Or 3000000 AIR a year.
I highly doubt that this reward rate will benefit Altair Ecosystem.
In absolute terms, it’s quite a lot but relatively speaking it’s still a small portion of the total supply.
Total supply - 470,000,000 AIR
Yearly collator reward - 3,000,000 AIR
That accounts for 0,6% going to collators which is actually not that much in comparison to other networks.
But there are two things to note - this number will decrease when market conditions get better (and that can take less than a year). Also, if this is the price for having enough decentralization and liveness for the parachain, it should be accepted - having operational cost in the range of $150-$200 is definitely not an exorbitant amount of money and we also give up on any profit in order to support the network in these hard times.
Based on the feedback of the collators and their running costs I believe collator rewards of around 20K-25K AIR are not unrealistic especially when we take the current price of AIR (around 1 Cent) into consideration.
We have already roughly 83% of the total supply of AIR in circulation which were distributed mainly for the airdrop and as CL-rewards.
Increasing the supply for collators, who are one of the core pillars of the ecosystem, should be considered
Good day Tjure07
I just expressed my personal opinion as community member. My opinion could be wrong, could be partial or in fully wrong.
I-m not a Team member, i-m not a Collator/Validator nor i`m Councillor.
So my feedback is just feedback.
If @pmensik or whenever else would like to change or propose something new:
Altair Collators Rewards changing
They could prepare the proposal with all changes, numbers and notes and Altair Councillors will discuss this with approval or refusal of this proposal.
If Pmensik or anyone else need help with the correct posting, feel free to contact me or @Rhano. I will be glad to help.
I`m also very glad that Pmensik started this discussion, because this is the first step that should be done in any governance process → Discussion on the Forum.
Great remark and that’s exactly how governance should look like. Identifiying a topic of interest, putting it into the forum to gather opinions and comments to use it as a temperature check for further proposals
honestly, I don’t feel it’s a good idea to run a mainnet node on Contabo (on any VPS) in general. We have some testnet and backup nodes on Contabo and we have experienced quite a lot of maintenance outages (even for a couple of hours), not to mention really bad performance. Substrate nodes are heavily dependent on CPU single-core performance in order to construct and propose blocks in a short period of time (500ms) and running the nodes on VPS means you can miss blocks quite often when it’s your turn to create one - I would be interested in what your monitoring says about it.
I agree with you, @pmensik, but as you pointed out the current Altair rewards don’t justify the costs. If we were rewarded $150-$200 / month for providing collator services, we would certainly consider different hosting solutions such as on a bare metal box with the specs you mentioned above.
Also there’s a thread on Polkassembly started with me and Paradoxx discussing the pricing of validators and collators so it might be useful here as well.
