CPU griefing over p2p

[faddat]

Background

The issue was discovered like this:

• because Larry sleeps near his validator and the fan kept him awake
• investigated by Jorge from Stargaze, who thought it was just a stargaze issue
• patched by Jorge from Stargaze
• seen on the Cosmos hub by Jorge from Stargaze
• discussed on twitter by Zdeadex from Stakelab, who was downtime-trolling Stakefish (deservedly, imo)
• Seen by Notional team members
• Long patched our Gaia validator and cpu usage immediately dropped
• Notional and Informal investigated, and in the end, the root cause was the sensors from Observatory.zone -- Clemens from Cryptocrew put us in touch with Rockaway, the makers of Observatory

All tendermint networks are likely effected.

Suggested Design

Use github.com/informalsystems/tendermint v0.34.26

Acceptance Criteria

Nodes should work without any functional difference whatsoever. If the operator has been experiencing higher than normal cpu usage, they should see that drop.

Closed by:

• 4267

[ValarDragon]

Closing as details of this don't need to be in open github issues, beyond apply security update.