I’m working with HardHat to deploy smart contracts on the Polygon network and running into a frustrating issue. Everything works fine most of the time, but I keep getting a “ProviderError: invalid sender” error when I try to switch between different networks.
Here’s what happens: I’ll be deploying contracts and creating tokens on one network just fine. But as soon as I change my configuration to use a different RPC endpoint, the error shows up. For instance, when I switched my hardhat.config.js from the main network to use rpc-mumbai.maticvigil.com, deployments started failing with this invalid sender message.
The weird part is that changing to matic-mumbai.chainstacklabs.com fixes everything and contracts deploy without problems again.
Has anyone else experienced this kind of network switching issue with HardHat and Polygon? Is this a common problem or am I missing something in my configuration?
Oh wow, this sounds super familiar! I’ve been dealing with the same headaches on Polygon deployments lately. The RPC endpoint switching is such a pain.
Quick question - are you using the same private key/account across both endpoints? I’ve noticed some RPC providers get really picky about account nonces or gas estimation when you jump between them.
Also, when you switch from maticvigil to chainstacklabs, do you clear cached data or just swap the URL? Wonder if there’s some state getting stuck that’s messing up sender validation.
Have you tried adding specific network configs like chainId or gasPrice when defining networks? Different RPC providers expect different parameters and Hardhat might not handle transitions smoothly.
What’s your wallet setup look like in the config? Using mnemonic or direct private keys?
This is usually an RPC provider issue, not your config. I hit the same problem with Mumbai testnet deployments last year. Different RPC endpoints keep different internal states for account nonces and pending transactions. When you switch from maticvigil to chainstacklabs, you’re moving to a provider with cleaner state. Explicitly resetting the provider connection in your hardhat config should fix this. Try adding accounts: { initialIndex: 0, count: 1 } to force account reinitialization. Also check if your deployment scripts are waiting for transaction confirmations before moving to the next step. Some RPC endpoints are more aggressive about rejecting transactions from invalid states.
yeah, I’ve hit this with polygon RPCs too. add a 2-sec delay between network switches - await new Promise(resolve => setTimeout(resolve, 2000)) before you deploy. some RPC endpoints are slow to sync state and mess up sender validation. also double-check your nonce isn’t getting reset - that’s usually what breaks it.