I think uptime matters more than pool fees for large mining farms

[hashradar]

I’ve been looking at different Bitcoin mining pools recently and something caught my attention.

Whenever miners compare pools, the discussion almost always ends up being about fees. People spend a lot of time arguing over the difference between 0.5%, 1% or 2%, but I rarely see anyone talking about uptime, rejected shares or stratum stability.

Maybe I’m looking at this the wrong way, but if you’re running something like 10 PH/s, wouldn’t a few hours of downtime cost more than the fee difference between most pools?

For those running larger farms, what would make you leave your current pool first: higher fees or reliability issues? 

I’d be interested to hear some real-world experiences because sometimes it feels like miners spend more time comparing fee percentages than comparing the things that can actually stop a farm from earning. 

[NotFuzzyWarm]

For any decent pool, downtime is not an issue. Most have multiple nodes around the globe so just set the miners to automatically fail over to one of them if a node goes down. I use Kanopool and since it started in 2014 it has had ZERO downtime. Very rarely one of its nearly dozen nodes around the globe might have an issue but Kano very quickly resolves the problem and in the meantime my miners just switch to a different node until my main one is back online.

I should add that if you have a few hundred PH or more (no rentals) at no cost to you Kano will setup a local private node connected to the main servers for you. Rented hash is only allowed on the Solo side of the pool.

As others have said, uptime of the miners themselves is by far the main issue faced by both home and large industrial miners.

[FP91G]

I’ve been looking at different Bitcoin mining pools recently and something caught my attention.

Whenever miners compare pools, the discussion almost always ends up being about fees. People spend a lot of time arguing over the difference between 0.5%, 1% or 2%, but I rarely see anyone talking about uptime, rejected shares or stratum stability.

Maybe I’m looking at this the wrong way, but if you’re running something like 10 PH/s, wouldn’t a few hours of downtime cost more than the fee difference between most pools?

For those running larger farms, what would make you leave your current pool first: higher fees or reliability issues? 

I’d be interested to hear some real-world experiences because sometimes it feels like miners spend more time comparing fee percentages than comparing the things that can actually stop a farm from earning. 

Some large miners make their own pools and do not care about commission problems.
Other miners get involved with such pools and also receive favorable commission conditions.
A large miner has a lower commission, a lower price for electricity, a good discount on ASICs with favorable installment plans.

[hashradar]

For any decent pool, downtime is not an issue. Most have multiple nodes around the globe so just set the miners to automatically fail over to one of them if a node goes down. I use Kanopool and since it started in 2014 it has had ZERO downtime. Very rarely one of its nearly dozen nodes around the globe might have an issue but Kano very quickly resolves the problem and in the meantime my miners just switch to a different node until my main one is back online.

I should add that if you have a few hundred PH or more (no rentals) at no cost to you Kano will setup a local private node connected to the main servers for you. Rented hash is only allowed on the Solo side of the pool.

As others have said, uptime of the miners themselves is by far the main issue faced by both home and large industrial miners.

That's probably true for established pools.

What I find interesting is that miner downtime is usually very visible, while pool-related losses can be much harder to spot. A miner going offline gets noticed immediately, but things like higher rejected shares, latency issues or slightly lower profitability can go unnoticed for quite a while.

In the end both affect revenue, but one is a lot easier to measure than the other.

Out of curiosity, how do you personally tell when a pool is underperforming? Do you mainly look at rejected shares, or do you track profitability over longer periods as well? 

[hashradar]

I’ve been looking at different Bitcoin mining pools recently and something caught my attention.

Whenever miners compare pools, the discussion almost always ends up being about fees. People spend a lot of time arguing over the difference between 0.5%, 1% or 2%, but I rarely see anyone talking about uptime, rejected shares or stratum stability.

Maybe I’m looking at this the wrong way, but if you’re running something like 10 PH/s, wouldn’t a few hours of downtime cost more than the fee difference between most pools?

For those running larger farms, what would make you leave your current pool first: higher fees or reliability issues? 

I’d be interested to hear some real-world experiences because sometimes it feels like miners spend more time comparing fee percentages than comparing the things that can actually stop a farm from earning. 

Some large miners make their own pools and do not care about commission problems.
Other miners get involved with such pools and also receive favorable commission conditions.
A large miner has a lower commission, a lower price for electricity, a good discount on ASICs with favorable installment plans.

That's a good point.

Large miners definitely have advantages when it comes to electricity costs, hardware pricing and pool terms.

At the same time, I don't think fees become irrelevant. Even a small difference can add up when you're running serious hashrate year after year.

My guess is that large operators look at the whole picture rather than any single metric. A slightly lower fee won't help much if the pool has other issues, but all else being equal, lower fees still mean more BTC kept by the miner.

[FP91G]

I’ve been looking at different Bitcoin mining pools recently and something caught my attention.

Whenever miners compare pools, the discussion almost always ends up being about fees. People spend a lot of time arguing over the difference between 0.5%, 1% or 2%, but I rarely see anyone talking about uptime, rejected shares or stratum stability.

Maybe I’m looking at this the wrong way, but if you’re running something like 10 PH/s, wouldn’t a few hours of downtime cost more than the fee difference between most pools?

For those running larger farms, what would make you leave your current pool first: higher fees or reliability issues? 

I’d be interested to hear some real-world experiences because sometimes it feels like miners spend more time comparing fee percentages than comparing the things that can actually stop a farm from earning. 

Some large miners make their own pools and do not care about commission problems.
Other miners get involved with such pools and also receive favorable commission conditions.
A large miner has a lower commission, a lower price for electricity, a good discount on ASICs with favorable installment plans.

That's a good point.

Large miners definitely have advantages when it comes to electricity costs, hardware pricing and pool terms.

At the same time, I don't think fees become irrelevant. Even a small difference can add up when you're running serious hashrate year after year.

My guess is that large operators look at the whole picture rather than any single metric. A slightly lower fee won't help much if the pool has other issues, but all else being equal, lower fees still mean more BTC kept by the miner.

I used to be able to outperform even large miners in terms of profits because I had cheap electricity and was working alone. Taxes were 0%.
Now my electricity is 5 cents, and I don't have any lucrative discounts on equipment. Commissions are out of the question. Furthermore, my operations may be illegal; the government is cracking down on private miners.
Ultimately, I'm losing to large miners on every level.

[hashradar]

I used to be able to outperform even large miners in terms of profits because I had cheap electricity and was working alone. Taxes were 0%.
Now my electricity is 5 cents, and I don't have any lucrative discounts on equipment. Commissions are out of the question. Furthermore, my operations may be illegal; the government is cracking down on private miners.
Ultimately, I'm losing to large miners on every level.

I can see where you’re coming from.  

It feels like scale is becoming more important every year. Large operations can negotiate better electricity rates, get better hardware pricing and often have access to terms that smaller miners simply can’t get.

Mining is increasingly looking like a real business rather than a hobby. That said, regardless of size, choosing the right pool still matters. Even if you can’t compete with a large farm on electricity costs, you can still avoid giving away more BTC than necessary through poor pool selection.  :-

[FP91G]

I used to be able to outperform even large miners in terms of profits because I had cheap electricity and was working alone. Taxes were 0%.
Now my electricity is 5 cents, and I don't have any lucrative discounts on equipment. Commissions are out of the question. Furthermore, my operations may be illegal; the government is cracking down on private miners.
Ultimately, I'm losing to large miners on every level.

I can see where you’re coming from.  

It feels like scale is becoming more important every year. Large operations can negotiate better electricity rates, get better hardware pricing and often have access to terms that smaller miners simply can’t get.

Mining is increasingly looking like a real business rather than a hobby. That said, regardless of size, choosing the right pool still matters. Even if you can’t compete with a large farm on electricity costs, you can still avoid giving away more BTC than necessary through poor pool selection.  :-

In Russia, miners don't use Foundry USA due to sanctions.
The classic three are ViaBTC, AntPool, and F2Pool. There are also local solutions, but it's up to each individual to decide which to use if they're mining unofficially.
I don't think the pool choice will significantly impact profits with 5 ASICs; in this case, it's better to choose a more stable pool with a lower ping.

[hashradar]

Good points from both of you, and I think between you you've actually answered the original question.

NotFuzzyWarm is right that for any serious pool, downtime basically isn't a thing anymore. The part I'd add is that you no longer have to take the operator's word for it. There's independent monitoring now that hits each pool's stratum endpoint every 5 minutes with a real mining.subscribe / mining.authorize handshake, so it's measuring "is the pool actually handing out work", not just "is the website up". Pull it up over 90 days and the big names all sit right at 100% - Foundry, AntPool, F2Pool, ViaBTC. A zero downtime since 2014 claim is great, but it's nicer to see it verified from outside.

What spreads more than uptime is ping, which ties straight into FP91G's reject-rate point: F2Pool ~110ms, ViaBTC ~113, AntPool ~160, Foundry ~400, and a couple of otherwise-fine pools sitting over 1000ms. That latency is invisible in the fee number but shows up in your rejects.

Which is the useful twist: once you filter for reliable pools, they're nearly all ~100% uptime, so uptime stops being the thing that separates them. What's left is net output after fees. Same FPPS model, a 4% pool vs a sub-1% pool is a few percent of your BTC every single day, and at 10 PH that compounds into real money over a year.

That's the split I track  : https://hashradar.live/uptime for reliability and ping, and the main board for net sat/PH/day. Right now the net-output leader there is actually a low-fee pool - Headframe at 0.9% FPPS is ahead of the 4% majors on sat/PH/day. For a large farm where every credible pool already clears the uptime bar, that fee delta is basically the whole decision.

[FP91G]

Good points from both of you, and I think between you you've actually answered the original question.

NotFuzzyWarm is right that for any serious pool, downtime basically isn't a thing anymore. The part I'd add is that you no longer have to take the operator's word for it. There's independent monitoring now that hits each pool's stratum endpoint every 5 minutes with a real mining.subscribe / mining.authorize handshake, so it's measuring "is the pool actually handing out work", not just "is the website up". Pull it up over 90 days and the big names all sit right at 100% - Foundry, AntPool, F2Pool, ViaBTC. A zero downtime since 2014 claim is great, but it's nicer to see it verified from outside.

What spreads more than uptime is ping, which ties straight into FP91G's reject-rate point: F2Pool ~110ms, ViaBTC ~113, AntPool ~160, Foundry ~400, and a couple of otherwise-fine pools sitting over 1000ms. That latency is invisible in the fee number but shows up in your rejects.

Which is the useful twist: once you filter for reliable pools, they're nearly all ~100% uptime, so uptime stops being the thing that separates them. What's left is net output after fees. Same FPPS model, a 4% pool vs a sub-1% pool is a few percent of your BTC every single day, and at 10 PH that compounds into real money over a year.

That's the split I track  : https://hashradar.live/uptime for reliability and ping, and the main board for net sat/PH/day. Right now the net-output leader there is actually a low-fee pool - Headframe at 0.9% FPPS is ahead of the 4% majors on sat/PH/day. For a large farm where every credible pool already clears the uptime bar, that fee delta is basically the whole decision.

https://hashradar.live/uptime
Pool Uptime Monitor
Independent monitoring of Bitcoin mining pool availability. Stratum, website, and dashboard endpoints checked every 5 minutes from a single location (Contabo VPS, Germany).
This is a good resource, but it only provides useful data for Germany. I don't think there are any small private miners in Germany because electricity for mining is very expensive there.
Germany has very strict environmental regulations, and the share of mining in Germany is less than 0.1-0.2% of the global hashrate.