After the Kusama active validator set was reduced from 1,000 to 700 validators, the remaining validators are already benefiting from roughly 30% higher rewards. That should have been more than enough.
Instead, some validators have chosen to exploit their nominators by setting outrageous commission rates of 50%, 70%, and even 85%. These commissions have nothing to do with covering operating costs, they are simply extracting as much value as possible from the people who trust them.
If a validator believes taking the vast majority of a nominator’s rewards is acceptable, they do not deserve the community’s support.
Please check your nominations, review validator commissions regularly, and stop nominating validators who abuse their position.
I also want to ask everyone with a strong presence on X, Telegram, Discord, Reddit, and other community channels to spread this message. Many nominators have no idea their rewards are being drained by these excessive commissions.
The best way to stop this behavior is simple: inform the community and remove support from validators who exploit their nominators. A healthy Kusama ecosystem depends on transparency, accountability, and fair competition—not on rewarding opportunistic behavior.
You can check their Kusama commission rates here, use the filters to display validators charging more than 20%, and remove them from your list. https://www.dozenodes.com/kusama
Right now anything at 40%-50%+ commission should be considered break-even (depending on company or operator tax rate) or more accurately anything under that rate is a subsidy for Kusama. A 40% should generate something between 150$ to 180$ (USD). Current data center prices have skyrocketed and KSM has plummeted in value.
Current market for the cheapest Hetzner machines (heavily discouraged due to policies) with minimal specs (remember to search for NVMe as other posts basically only looked for CPU and not even +Zen3) is 150€ the next one is already at 230€. Auctions don’t offer NVMes atm. OVH has them at 160$ (tho you’ll need to up the bandwidth from the minimum to 50TB most likely, or you pay them extra by the end of the month) Scaleway has them at 179€ so you get the idea. Add DEX/CEX conversion rates, tax and other expenses then Kusama has already become a “mining at a loss” enterprise for many with fees under 40%. Also keep an eye on a further increase as it’s about to become a 4TB disk needed blockchain.
So propanganda debooonked and honestly this line of thinking should be discouraged as it came from a marketing campaign looking to demonize validators for all the wrong and currently detrimental reasons.
Q&A:
Q: but I found this one super cheap machine at Ghettostein by Hetzner.
A: 700 current machines won’t be able to use that particular one or change nimbly as these prices are obviously outliers at the moment.
Q: The market will find a solution. Free market is good. 0% commissions are good.
A: That only has led to a centralization and not by exchanges, VC, etc. Check the other posts about the Nakamoto numbers for more info. Or any other chain that has done that.
The numbers you’re using are misleading because they’re based on the assumption that every Kusama validator needs its own dedicated 64 GB server. That’s simply not true.
The official Kusama documentation recommends 32 GB of RAM, and in practice many operators run validators perfectly fine with 16 GB, depending on their setup. In other words, a single server can comfortably host multiple validator nodes, dramatically reducing the actual infrastructure cost per validator.
Why delegate to a validator with 50% or 85% commission when there are experienced operators providing the same reliability, uptime, and performance for 15% commission?
Don’t blindly believe everything some validators tell you. Take a look at the numbers and do your own research.
There are plenty of reliable, experienced operators providing the same level of service for 15% commission. Paying 50% or even 85% doesn’t buy you better security, better uptime, or better performance, it just means giving away a much larger share of your rewards.
It’s not misleading at all and the numbers still stand, these are the real earnings per node something you can check onchain on Subscan.
75$ to 90$ at 20%
150$ to 180$ at 40%
300$ to 360$ at 80%
etc. Surely some variance here and there but that’s pretty much it. Not sure if the “greed” narrative really holds up when the max you could extract with a 100% commission is less than a MCds burger flipper salary in a third world country
And expenses are only following the minimum requirements so it’s not misleading at all, all the companies prices are public too. FYI DC prices at the moment are never about the RAM, it’s about the disk, just check the sites previously mentioned. Hetzner even starts at 64GB most of the time but with tiny disks where the blockchain is not going to fit or slow disks where the blockchain will never sync.
Sure cutting corners is what everyone is doing and you are confirming that many are already operating under the minimum requirements. But going with that thread we can also prove that on Hetzner EPYC 9000s/7000s with +32 cores and +8TB/10TB of NVMe go for 700€-800€ in this very moment. You will find them in the +300€+400€ range but they will run out of disk for sure as they have smaller disks than that so you only get to run 2 at the time in such machines. Remember when POS DOG went down altogether? Also, 4TB disks coming soon .
These are the current prices pretty much without any significant variation.
“Just use one core for Kusama” well, we are already seeing degradation on other machines suggesting that that might be the case which in turn cause degradation in the group validating so probably we are already seeing the effects of what lowering “dramatically” compute can cause.
You can drag losses for some time, have contracts done in the past (two years ago would have been optimal) specially with longer contracts and single racks absorbing 20% of Kusama. See the answer and threads about “Q: The market will find a solution. Free market is good. 0% commissions are good.:”
This reminds me of the exact same discussion we had around the IBP Bounty cost model. Back then, the operating costs were calculated using AWS pricing, yet almost none of the operators were actually running on AWS. They were using other providers that were just as reliable but significantly cheaper.
I see the same issue here. Unless you’re running archive nodes, I don’t understand where the storage concern comes from. Community reports consistently place a ParityDB database with pruning enabled at around 200–400 GB, even after long periods of operation. That’s nowhere near requiring a dedicated server for a single Kusama validator.
I can confidently say that a single server with 64 GB RAM and 2 TB of NVMe storage can comfortably run 3 or 4 pruned Kusama validator nodes.
So no, I don’t think these infrastructure costs justify the 50% or even 85% commissions charged by some operators. It feels like some are trying to convince nominators that running a validator is far more expensive than it actually is.
There’s nothing wrong with earning a fair profit, but charging excessive commissions while portraying validator operations as prohibitively expensive is misleading. Nominators should do their own research and remember that there are many experienced, reliable operators providing the exact same service for 15% commission.
According to your own estimates, a validator earns around $300–360/month with an 80% commission.
Now let’s assume that server runs 3 pruned Kusama validator nodes, which is perfectly realistic. That would generate $900–1,080/month in revenue, while the infrastructure cost would still be only $155/month. In other words, the operator would be making approximately $745–925/month before accounting for any additional operating expenses.
Even with just 2 validators, the revenue would be $600–720/month, leaving $445–565/month after paying for the server.
The bigger issue is that Kusama has been losing attractiveness for a while. The token has underperformed, inflation remains high, and many validators sell their KSM rewards immediately to cover infrastructure costs while holding onto their DOT because they see more long-term potential there.
That said, increasing the minimum commission certainly didn’t help. It reduced competition on fees and forced nominators to accept higher costs even though many operators could still provide the same level of service for 15%.
@SAXEMBERG I don’t think the situation is that bad as you say, if that were the case there wouldn’t be 1044 validators on the waiting list. Even KSM had to reduce the number of validators because there were simply too much of them. When the minimum comission was raised I saw many nominators leaving since it was not worth it anymore and KSM value plummeted. We should do something to help attract nominators back. Maybe lowering min comission to 0% be a good start.
Reality is validators are desesperate to validate on Kusama and the waiting list is a proof of that.
Also thanks for confirming that RAM is not the issue for providers
Sure you can lower the bar for Kusama but that actually means running a validator with 2-3 physical cores which was said almost as a joke here:
Archive already needs 4TB.
Wow, this is so painful to see . No wonder Kusama is having so many stability issues now. Also you shouldn’t use ParityDB, check element messages about this topic. Again, this is taken from the minimum requirements:
There is so much hypothetical baloney here, there are only 3 machines that could achieve this in the active set assuming they belong to the same operator/company and they seem they are not. All the rest are well below that percentage. The example was just to explain how little it’d be if even if it was set at 80% . Now this is focusing on hypotetical non-actual events.
In any case, let’s not push for a further degradation of the infrastructure of Polkadot and Kusama. We already know we can shrink infra beyond current capabilities but actively advocating for it when the real issue is the price of the token set by the market is honestly mindblowing. In any case, most of this thread has been mostly debunked as there are no yatches bought with KSM inflation.
I’ve explained my reasoning and even provided a real-world example from an operator I work with who runs multiple pruned Kusama validator nodes on the same server. If you disagree, that’s perfectly fine, but simply calling it “baloney” doesn’t address any of the points I’ve made.
There is no reason for nominators to delegate to validators charging 50% or even 85% commission when there are operators with A+ performance providing exactly the same service for 15% commission.
If you choose to run archive nodes that require 4 TB or more of storage, then that’s a completely different use case and, naturally, you have a storage problem.
Dear validators, verify your expenses and income with receipts and detailed statements of income and expenses.
The mere fact that this has never been done is suspicious. All we hear is you whining, yet you’ve been shamelessly fleecing the nominators for years. Your role model is W3F, which does the same thing.
Ironically, you would have earned far more if you had reduced the inflation rate and focused on limiting the oversupply, but you’re apparently too stupid for that.
Enjoy your downfall—no one will take you seriously in the future, and no one will even want to buy a hot dog from you, let alone elect you as validators.
.
. Now this is focusing on hypotetical non-actual events.