Please login or register.

Testnet Hard Fork: 23rd November, 2015, block 624634

We're testing next year's mainnet fork (MRL1, MRL4, block time change) in the next ~5 hours, hold thumbs that it doesn't explode into a million glittery pieces:)

If you want to participate you can run a testnet node, either from the 0.8.8.6 binaries from the site (to see how badly it breaks) or compiled from master (to see if you transition cleanly to the new fork). You probably won't have enough time to sync up from scratch, though.

I'll update this thread after the fork!

Replies: 4
8xmr posted 8 years ago Weight: -17 | Link [ - ]

Congratulations!

fluffypony posted 8 years ago Replies: 1 | Weight: -75 | Link [ - ]

Ok - testnet fork succeeded, there were a couple of hairy moments where we thought it had gone wonky but it was alright.

One thing we did notice is that old nodes are connecting to new nodes, requesting blocks, can't parse them, and then aggressively reconnecting. We're going to add functionality to graylist bad nodes for a longer period, maybe a week or something.

Reply to: fluffypony
rocco edited 8 years ago Weight: -75 | Link [ - ]

Good to hear and happy you even found some bugs, sounds like a successful test to me!

btw yeah i have a node like this. failed to parse block from blob. coool :D

smooth posted 8 years ago Weight: -75 | Link [ - ]

"One thing we did notice is that old nodes are connecting to new nodes, requesting blocks, can't parse them, and then aggressively reconnecting"

Seems less likely to be an issue with the actual fork procedure when it doesn't activate for several months. Especially given the age of 0.8.8.6 and enormous performance improvements, I expect nearly everyone to have upgraded (unlike on testnet). For future forks when the upgrade is less compelling it may be a bigger issue.