The upgrade is scheduled for block 8576398
, which should be about 17:30 UTC on 8th June 2023.
These instructions assume you are running Cosmovisor. Most mainnet validators are running Cosmovisor, and a setup guide can be found here.
NOTE: Cosmovisor will preform a full backup unless UNSAFE_SKIP_BACKUP=true
is set as an environment variable.
We recommend using go version 1.20.4 which has the latest security updates.
# get the new version (run inside the repo)
git fetch origin --tags
git checkout v10.0.1
make build && make install
# check the version - should be v10.0.1
$HOME/go/bin/starsd version --long
> name: stargaze
> server_name: starsd
> version: 10.0.1
> commit: 8ca01207d10ce0aff5b2b984635e6495d5c42198
> build_tags: netgo,ledger
> go: go version go1.20 linux/amd64
# make a dir if you haven't
mkdir -p $DAEMON_HOME/cosmovisor/upgrades/v10/bin
# if you are using cosmovisor you then need to copy this new binary
cp /home/<your-user>/go/bin/starsd $DAEMON_HOME/cosmovisor/upgrades/v10/bin
# find out what version you are about to run - should be v10.0.1
$DAEMON_HOME/cosmovisor/upgrades/v10/bin/starsd version
If you are not using Cosmovisor, then the chain will halt at the target height and you can manually switch over.
If you are using v10.0.0 instead of v10.0.1, after the v10 upgrade automatically halts the chain, remove the wasm cache located in rm -rf ~/.starsd/wasm/wasm/cache
and start the node with the new version.
If you see the following error stop the node remove the cache and start the node again.
fatal error: unexpected signal during runtime execution
May 30 20:53:59 s631 cosmovisor[115240]: [signal SIGSEGV: segmentation violation code=0x1 addr=0x7fd315131000 pc=0x7fd37c2d9cd0]