We have a live testnet chain where we planned a runtime upgrade (RU). Before performing RU in the live chain we performed fork-off of that chain to get all the states same as it is and started a local chain with that and performed the runtime upgrade there in which the RU and the storage migration was successful. But when we tried it on the live testnet RU is successful but the storage migration didn't happen correctly. The storage which we targeted to migrate got completely erased. What could be the possible reasons?

EDIT 1 : We found the root cause. In the live chain the pallet storage version is 0 but while we start the fork-off with the same state pallet storage version seems to start from 1 for every pallet that we use. That's the reason the migration has been skipped in the live testnet. Why the storage version is different in forkoff?

  • Need more detail. E.G. steps of fork-off and launch.
    – AurevoirXavier
    Commented Sep 1, 2023 at 1:55
  • @AurevoirXavier edited my question. pls check Commented Sep 1, 2023 at 3:23
  • You mean fork-off-substrate? Please do use the proper tooling and not some known-to-not-work things. The advised way is to use the try-runtime-cli. Commented Sep 1, 2023 at 11:40


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.