I was scolded on Reddit for suggesting farmers download a copy of the DB from a site I learned about on here, and have repeated to others. Screenshot of the admonition below.
2 Likes
What was said is true. You should not download a copy of someone else’s database and use it. There is no way to know if it has been tampered with. I’m sure that any thread on here or anywhere else that offers a download will be full of responses saying not to do it and if you do it is at your own risk.
The only way to ensure you get a complete and accurate copy of the database is to let the node pull it from other nodes and confirm each block as it does it. This ensures the accuracy of the blockchain. yes, it is slow, but that is the only way … currently.
3 Likes
The idea that the only valid source of info in the Chia community comes from the top down bugs me to no end. That’s not how decentralized systems work, and it stifles innovation.
This reply pretty accurately explains the risk of bootstrap downloading the blockchain to speed up syncing: Elysium Pool is releasing a blockchain bootstrap download service - #5 by elysiumpool
Regardless of whether or not you support bootstrap downloading the blockchain, which is provably safe, I would suggest that members of the community call upon the devs to come up with a much better system for syncing the blockchain, as the current system takes days on modest hardware.
Nearly every other established cryptocurrency (bitcoin, litecoin, ethereum, etc.) has a preferred procedure for downloading a blockchain bootstrap.
4 Likes
The problem is, and I don’t see it is likely to happen personally.
If many many ppl were running a altered db, nothing would get found when syncing as others would also have that altered db so finding peers would not be an issue.
This could then cause a chain split if enough ppl had that wrong db.
I think in order for it to be a problem many would have to run that altered db at once though, something I think unlikely, and why I dont think it would happen.
2 Likes
Thanks all. I had presumed the bootstrap was somehow sanctioned by Chia. If not, then an appropriate workaround would be very useful.
I couldn’t find my earlier posts, perhaps they were appropriately removed?
I think you’re misunderstanding how blockchains work. The longest blockchain with the most blocks mined is synced as the correct blockchain.
When blocks (which contain metadata and transactions) are synced/transferred between nodes, each block is verified with following blocks.
If the situation you’re describing was possible, don’t you think someone would try attacking the network by trying to sync a malicious altered blockchain to other nodes. No one has done this because it’s not possible, regardless of how the altered blocks appeared (whether by a malicious actor hosting their own node, or by a malicious actor spreading them via a blockchain boostrap snapshot).
3 Likes
What both of you are saying is correct. But you just aren’t seeing what Bones is saying. It is far fetched (like he said) but is correct. Let’s say 2000 people downloaded a modified database. This database was modified by me so that it shows I was awarded 7000 XCH over the last 5 months. All 2000 of these people have this DB loaded on in their nodes and they are live along with myself. That is plenty of people to validate that this version of the blockchain is legit and any new and old transactions would be verified and completed. This chain has now split from the original chain. But I now have 7000 XCH.
Would this happen. Probably not. But can it, absolutely. Because as you stated, that is how blockchains work.
2 Likes
How could you do anything with that 7000 xch? No nodes would accept your transactions
2 Likes
As long as the nodes are connected to the network though, they will see that the fake blockchain is fake as the wallet module verifies the blocks in the blockchain, and sees that the merkle hash tree does not match.
I would love for someone to employ this small-scale and see that this is not a viable attack vector.
2 Likes
That’s why it’s called a chain split.
The xch created in the above example would be completely valid, but only on that split chain.
It’s not only possible but has happened.
Many years ago on bct, botnet owners used to openly threaten and attack small projects they didn’t like and used this tactic to destroy them.
The attack worked because if the split chain gets large enough, new incoming nodes connect to more peers in that chain , syncing from that chain , this makes the original chain redundant.
1 Like
What you’re describing is actually a 51% attack. These attacks are completely unrelated to the discussion here.
1 Like
What am I to do with my 8500 XCH,just kidding
1 Like
Well, we will agree to disagree then.
A chain split can and has happened long before that chain reaches 51% in order to make the initial chain redundant.
Of course the main chain is safe up to that point.
But anyone who has downloaded an altered db and is farming it has wasted their time and farmed useless xch on a useless chain.
Ask yourself if there’s no risk, why won’t the devs support it?
1 Like
if there’s no risk
I never said there was no risk. I simply suggested that anyone with half a brain should be able to see that the risk is so minuscule that it need not be weighed.
Assuming you are connected to even one valid node as a farmer, you are guaranteed to be farming on the real chain.
Isn’t there a checksum in the DB? And wouldn’t that be extremely hard to modify successfully to benefit something? And then get it to enough people to effectively make it the source of truth?
Why do I see pitchforks? Uhhh, Okay I’m going to run now!
1 Like
You claimed
But failed to see that the chain split can happen even without 51%…
So was completely relevant to the discussion.
Wrong, if your connected to more nodes on the altered chain, you will disallow the real chain and fall in with altered chain.
1 Like
Do you have a source or example of this? What you’re describing isn’t consistent with what makes Chia and all blockchain cryptocurrencies strong (i.e., proof of work or proof of storage allows nodes to trustlessly determine which chain is correct).
It was commonplace years ago on bitcointalk, but I’m not searching 8 year old threads to find them.
Eth chain nearly split just last year as ppl were running old software.
All that power in pow or storage in pos is purely trying to mint the next block, it has absolutely no concern which chain its minting blocks on.
Which chain its minting blocks on is purely determined by its db and node concensus, so if your connected to more nodes on a particular chain, you will re org your blocks to that chain and mint blocks on that chain.
Is there any reason to believe this is occurring now, on the Chia blockchain?