How to prepare your node for v3 upgrade
This guide includes all the steps you need to get ready for the v3 upgrade in Bostrom.
Hereby we assume that you run your node in docker with the cyber "official" image, otherwise, this guide will not suit you.
Step 1: Pull a new docker image
The new cyber binary is wrapped in the new docker image, so we have to pull it first:
docker pull cyberd/bostrom:v3.0.0-cuda11.4
It may take a few minutes. When complete, check that you have it:
docker images
Step 2: Replace the docker image for the Bostrom container
At this stage you will have to operate quickly to reduce your node downtime, so please read all the sub-steps first!
Stop your docker container with Bostrom:
docker stop bostrom
Remove Bostrom container (do not worry, files will stay intact):
docker rm bostrom
Create Bostrom container from new image (note: if you have custom ports set, please adjust accordingly):
Check logs. Block production should be restored after startup:
Step 3: Verify that everything is OK
After the container launch, the new cyber binary should appear inside the $HOME/.cyber/cosmovisor/ folder. Check that you have a similar folder layout:
You should have a similar layout of files and folders:
You are done! What's next?
Now is a good time to go and vote for prop #41. If you haven't done this yet, report that you have completed preparation for the upgrade in our Hall of Fame and chill until the upgrade block 13078000, which should come around 12:00am UTC Wednesday 17.04.2024 (ping.pub timer).
What will happen at the block 13078000:
- the chain will halt with the message
CONSENSUS FAILURE!!! err="UPGRADE \"v3\" NEEDED at height: 13078000: - Cosmovisor inside the container will catch this event and change symlink inside
.cyber/cosmovisor/folder - node will start and as soon as 66%+ of voting power is online, blocks will resume
Crisis protocol
The process of upgrading should go pretty smoothly and should not require any direct action, but in case something goes wrong, cyber devs may consider certain ways to recover. Simply check your node at the time of the upgrade and consult the Hall of Fame chat for possible coordination.