Blockchain 💣
Opened this issue · 6 comments
Around 2 hours ago the graft blockchain experienced a (normal) orphan from simultaneously mined blocks, but then the orphaned chain refused to accept the main chain as valid and thus stayed permanently split. Spacepools (which found the orphan-to-be at height 307841), the Majanetwork pool node, my node that runs graft.observer, and at least one user's node in the Telegram channel all got stuck with spacepools' block and all refused to accept the main network chain that should have resolved the orphan.
This is not normal behaviour: the alt chain situation should have been resolved as soon as the next block was found, producing an orphan for spacepools which should have reorganized and continued normall. This did not happen: the affected nodes all considered the proper winning chain to be invalid.
I've uploaded a log from my graftnoded to https://jagerman.com/graft.log
In order to get it synced again I had to stop it, use graft-blockchain-import --pop-blocks 100, then start it again (the restart is at 2019-03-19 18:43:56.011 in the log file). 100 was probably excessive, but did the job. This sort of manual blockchain popping should never be necessary unless there is a serious network concensus bug or local hardware issue (but since this happened on multiple nodes on the network at once, we can rule that one).
I'm still getting alternative blcokchain when syncing do you have a list of recommended nodes?
I had some joy suing 45.76.62.202 but i don't want to leave running with 1 node.
I am currently on an alternative chain that split at block 308992 . How to I get back to the correct chain?
Block chain is currently whooped. Alternative is present and whacks everything. You would need to find a node thats on the proper one and sync from it. Use graftnoded-blockchain-import --pop-blocks 10000.
Still no response to this issue, I have attempted to sync the blockchain more than 4 or 5 times in the past month and not once have a managed to sync successfully, each time I have been forced to download the raw image of the chain, I am currently trying again and will post the graft.log
I have logged #284 for the issue I suspect I was facing for the above mentioned syncing attempts
Everyone Just remote node to the Dev pool. Its always on the correct height. You guys shouldn't be attempting to synch your node if you don't know what your doing! You should always rely on Remote nodes. If everyone remote nodes to the Dev pool everyone will be on the same height and this will not be an issue! Just open up your Gui to vew the transaction when you send to make sure everything is okay.
On a serious note. In the original post. Im not sure if spacepools the one that had the orphan has some synch issues... But for some reason his pool always seems to lag behind the rest of the network when synching. Ive seen all the pools swap to the new block and his pool lag for like 30 seconds before switching. That's a long ass time. Also why would the winning chain not be accepted? What in the codeing could affect this?