To all validators, just to be clear, this is a hard-fork upgrade. Kindly read the instructions for v2.2.0 clearly.
There will not be any voting before the validator upgrade. this upgrade is to enable certain features of the blockchain. to activate these features, the governance proposal will happen after the upgrade
As explained on TG, here are the issues I found (on testnet upgrade only):
My testnet validator appears INACTIVE in dhobyghaut-explorer
Its latest block is 3920765 (25/JUL/2022)
Its voting power still appears at value 138.
When trying to unjail, response is Error: rpc error: code = InvalidArgument desc = failed to execute message; message index: 0: validator not jailed; cannot be unjailed: invalid request
Looks like almost all testnet validators are now inactive…
@FrenchXCore there is a need for another round of upgrade. on testnet. So this time, both mainnet and testnet have to be upgraded. There was a fix of a bug from the last upgrade which is why testnet needs to be upgraded again. @ClaudioxBarros
@Fox_Coin@Aravan please refer to this Upgrade countdown timer can be found here. for all future upgrades, the countdown timer link will be the same.
@ClaudioxBarros you can perform any commands from another terminal through your validator node. just use the CLI and specify the --node flag and input the ip address of your node. Something like this example command
apologies to the validators who have already upgraded your nodes. we forgot to merge the latest commit into the release/v.2.2.x branch, so you will have to upgrade again.
if you run git log, you should be able to see the tag v2.2.1 inside
commit 5eef07630e89ad0bd786fb08fa0fb937e5c84d67 (HEAD → release/v2.2.x, tag: v2.2.1)
Your version is v2.1.1. The height of the last testnet upgrade was 3918000. You did not upgrade to v2.2.0 before this height. After the height, if there is an upgraded logic triggered by a transaction, the unupgraded node will fork.
Solution:
either you upgrade with v2.2.0 before this height after rolling back the data or
my recommendation clear out all the old data and use the snapshot
I successfully updated our mainnet nodes to v2.2.1 (release d5d5…).
However, I still have the same issue as specified in quotes on testnet node.
Here’s some more error log:
Jul 29 11:08:52 fxcored[46911]: panic: Failed to start consensus state: found signature from the same key
Jul 29 11:08:52 fxcored[46911]: 11:08AM INF found signature from the same key height=3920765 idx=26 module=consensus server=node sig={"block_id_flag":2,"signature":"u917oh36nEU7I01n6x7kT/W2bGpohh8jleN3GSbsejV>
Jul 29 11:08:52 fxcored[46911]: 11:08AM ERR error on catchup replay; proceeding to start state anyway err="cannot replay height 3920766. WAL does not contain #ENDHEIGHT for 3920765" module=consensus server=no>
J
in your log i saw line containing: Jul 29 11:08:52 fxcored[46911]: panic: Failed to start consensus state: found signature from the same key
the fix for this was given before
What should I do if the “panic: Failed to start consensus state: found signature from the same key” appears in log?
We should first check whether the priv_validator_key.json configured by the node is used by its node, if not, run the command fxcored config config.toml consensus.double_sign_check_height 0, to modify the double-sign check, and then restart the node
could u try whether it fix and screenshot any other error that persists
So, my testnet node won’t progress above block #3920765.
Block #3920766 still generates a commit is for a block we do not know about; set ProposalBlock=nil…
What did I do ?
I reset completely a new testnet node, and did not enter my validator information at that stage (using release/v2.2.x).
I downloaded 20220725 testnet snapshot and ran the node to full sync.
I started it and waited…
I don’t know what happened around that block time, but it seems tendermint is refusing to validate it, thus making it impossible to progress beyond that block #3920765: fxcored[70034]: 8:47AM ERR CONSENSUS FAILURE!!! err="+2/3 committed an invalid block: wrong Block.Header.AppHash. Expected 6C93556A367CAF87A7F53539D58E3E646CDDFC44C174460E5A3AEB2AC07E59CD, go>
It looks as if more than 2/3 of the validators did validate block #3920766 with a different AppHash.
Would it be possible to publish a new testnet snapshot so that I can restart my testnet validator above that stage ?
On the other hand, I think it would be interesting to dig in what happened exactly, because that would hurt badly if this was to happen on mainnet.
My testnet face the same issues, i fixed it by clearing out old data, using the 25072022 snapshot too, mine was docker so i did remove all images/containers and follow the upgrade steps
Alright…
I started a brand new server and release/v2.2.x node from scratch and tried to sync it from the latest 25/JUL/2022 snapshot, and I still get the same error at block #3920765.
I also tried to sync without snapshot from Block #0… However, something strange hit my eyes.
The first line below shows a testnet/v2.0.x
Its less than 2.5 hours to the upgrade height. If there is anyone who has yet to upgrade and needs some sort of tech support, please reply to this thread!