New answers tagged

0

I found my issue. We have our Geth run command inside of a configMap that gets loaded by the pod. I manually added the --rpc.gascap 0 line to the configMap, and forgot to add a \ after. The following settings were being omitted from our startup command. I think the one that was missing that was causing problems was not setting the --datadir /ethereum/ ...


1

If the smart contract sending or receiving doesn't emit an event, then there is no way to know about the transfers without getting the transaction traces


0

Hardhat and ganache too are test tools and do not really show the real life scenarios that can happen when deploying to a real node. the most apparent difference is latency when broadcasting transactions since in a test setup mining doesn't really exist. another thing is the kind of attacks that only appear when using the actual geth node, like frontRunning (...


0

First of all, when doing the initial sync, you should do a "fast" or "snap" sync, not a "full" sync. Fast and snap sync will take about 400Gb initially. When the initial sync is completed, the node will automatically switch to "full" sync. Following that, the node will increase about 40Gb / day as you observed. There ...


0

Blocks aren't pruned. They are stored in geth/chaindata/ancient directory and you can't delete them because your Geth is serving this information to other nodes that at any point in time would need to sync just like you. The only pruning that takes places is State pruning, old state objects are deleted from Merkle Trie as new blocks are added. The latest ...


0

Just to update this answer, there is a requirement parameter on the AsMessage call. you need to pass a base fee, though the transaction is a free on, place a big int 0 into the call as shown below: msgFrom, _ := tx.AsMessage(types.NewEIP2930Signer(tx.ChainId() ),big.NewInt(1) )


1

What you have is part of the solution. You also have to look at events from that transaction. So eth_getTransactionReceipt will return the logs (and other data that will be useful for you.). also look into eth_getTransactionByHash Topic 0 is the event signature. so you would have this tab info:


1

There is no simulation in Geth, all transactions go for real. It is just that locally you have your own chain, all transfers are executed on your chain and nobody sees them. You can not connect to Ethereum Main Net because the hash of your block 0 (genesis) is different from the Main Net, so you will never interfere with the Main Net. Simply, nobody will ...


1

It would create a fork, which is soon to be rejected by the rest of the nodes, thus you get desynced, yes. No, it's only for reading, for writing you need eth_sendTransaction. More on this here: https://eth.wiki/json-rpc/API I'm not sure how much debugging Geth does provide but this might help: https://github.com/trufflesuite/truffle-plugin-debugger. Simply ...


1

Truffle Plugin Debugger (formerly Truffle Teams) can help: https://github.com/trufflesuite/truffle-plugin-debugger


1

it is: --miner.gasprice if you are using geth 1.10.6 geth help for more info


1

OK, I got it. It's a geth bug. A new release is published to fix it. https://github.com/ethereum/go-ethereum/releases/tag/v1.10.6


-1

Following up with @tqd's answer, please put it in the "config" block of the genesis.json, not in the root. Wasted so much of my time.


1

Mist and Eth wallet have been decommissioned in 2019. https://avsa.medium.com/sunsetting-mist-da21c8e943d2 Use something else.


0

Something was wrong with the storage in my node. I pointed storage to a new directory and now "snapshot" appears to be working fine, it has synced over 80% in last 48 hrs. Geth outputs status at this point : INFO [07-22|12:50:15.193] State sync in progress synced=81.94% state=101.06GiB accounts=129,924,534@26.62GiB slots=361,649,239@72.61GiB codes=...


0

Connected peers are used to fetch and broadcast data. A1 : the connected peer count is not a big deal if you know the quality of the peer node. a 1 node with the best hardware is better than 10 nodes in china running on dual-core PCs. So your best bet is to have the best nodes connected. Also having 200 peers on your node when you are running on low-end ...


0

This is straight from the documentation: web3.eth.sendTransaction({from: '0x123...', data: '0x432...'}) .once('sending', function(payload){ ... }) .once('sent', function(payload){ ... }) .once('transactionHash', function(hash){ ... }) .once('receipt', function(receipt){ ... }) .on('confirmation', function(confNumber, receipt, latestBlockHash){ ... }) .on('...


0

I got this error immediately after I uninstalled MetaMask. It might be that MetaMask was providing the Web3 library and you may need to include that library directly into your project.


0

Of course you can use the same contract for all transactions. In fact it's meant yo be used this way. You only deploy it once and all transactions are run by the same contract. You need to clarify what do you mean by "my website address" because the way you have it written the "fee" will remain in the contract. You would need to still ...


1

Go to ERC725 root directory and do npm install because your error shows that openzeppelin files are not found so most probably you forgot to install dependencies.


1

This works assuming your code includes the IERC20 library (this also works with BEP20 tokens ofc, they're essentially the same). It's trivial to modify it so you can withdraw only a certain amount or to another address. function withdrawToken (address tokenAddress) public { IERC20 token = IERC20(tokenAddress); uint256 balance = token.balanceOf(...


0

Holy moly.... To fix it I had to restart it completely (deleting the data) and using: docker run -d -v /data/binance:/root --name binance-smart-chain-node -p 0.0.0.0:8545:8545 -p 0.0.0.0:8546:8546 -p 0.0.0.0:6060:6060 -p 30311:30311 -p 30311:30311/udp vlddm/binance-smart-chain-node:latest --syncmode full --cache 4096 --rpc --rpccorsdomain="*" --...


1

M H Swende (holiman) said: It's cross-chain replay protection, not cross-block or cross-fork-threshold or whatever. Reference: https://discord.com/channels/595666850260713488/745077610685661265/865664081767956490


0

What you have is an old Ethereum node that is completely out of sync. /keystore Warning danger danger! If you had files in /keystore, you should keep them. Those files are your encrypted private keys (decrypted with a password that you set). The private keys give you control of your addresses on the blockchain. Lose these files and you lose all the tokens on ...


1

I found what was the reason for this issue for my case. I was using Geth version 1.10.3-stable. So I looked for the new releases at go-ethereum github page that if new versions solving the issue like this. And there were two new stable releases. And 1.10.4-stable version said that Geth v1.10.4 is a feature release and adds compatibility with the upcoming ...


0

Turns out this was a default message which now has been removed https://github.com/sigp/lighthouse/issues/2452


1

Instead of running ps and piping it to grep, you can use pgrep geth. One-liner: kill -HUP $(pgrep geth). – Diti Aug 13 '17 at 12:36 I did not see the above as an answer, but as a response to one of the answers. I used kill -HUP $(pgrep geth) and it worked perfectly, so thank you Diti


7

Nothing prevents collusion From former Bitcoin Core developer Greg Maxwell https://old.reddit.com/r/Bitcoin/comments/ddddfl/question_on_the_vulnerability_of_bitcoin/f2g9e7b The attack would work just as well if there were 100 people each with an equal amount and a majority of them colluded to dishonestly override the result. Also, any mechanism that would ...


3

As Mikko mentioned, it would be super difficult for them to start cooperating in such a way. If their miners find out, they can just easily switch to an "honest" pool. But assuming that those miners somehow wouldn't find out and the pools managed to cooperate, then yes they could cause some havoc. If they for example caused a week's worth of reorgs,...


1

if there are lots of events, you need to query by block range, otherwise you will wait for hours without getting a response. In golang this would look like: filter := ethereum.FilterQuery{} filter.FromBlock = big.NewInt(block_num_from) filter.ToBlock = big.NewInt(block_num_to) topics := make([]common.Hash,0,1) signature := common.BytesToHash(event_signature) ...


1

GoEthereum does not expose this information directly. You need to implement your own mempool watcher that detects transactions that were later replaced by a higher fee.


1

Remote Sealing Verification Remote Sealing Verification verifies if the block has a valid seal according to the POW (Proof-of-work) algorithm. Proof-of-work Proof-of-work is the underlying algorithm that sets the difficulty and rules for the work miners do. Mining is the "work" itself. It's the act of adding valid blocks to the chain. This is ...


Top 50 recent answers are included