We have a situation where a node using warp sync correctly warps until the finalized block, then it imports blocks that are part of a fork, and never recovers from there. Here are the logs
in this case, 450385 was the finalized one, and while it succeeds warping there, then it imports 450386, 450387, 450388, 450389 from a reverted fork. What happens next is that it tries to import the legit 450390, but since it does not have the parent, it gets stuck
2025-11-24 09:06:06.767 WARN tokio-runtime-worker sync: [Container-2005] 💔 Error importing block 0x014c9158dac97aa1a94623e88f6ebb83b419821c05ac34a6873c7ff3c4fd60ef: block has an unknown parent
After that, it gets stuck in idle, with the wrong best block
INFO tokio-runtime-worker substrate: [Container-2005] 💤 Idle (3 peers), best: #450389 (0xd936…7637), finalized #450385 (0xbcd7…9caf), ⬇ 108.9kiB/s ⬆ 6.6kiB/s
This happened while in stable2506, and it does not get fixed if restarted
We have a situation where a node using warp sync correctly warps until the finalized block, then it imports blocks that are part of a fork, and never recovers from there. Here are the logs
in this case, 450385 was the finalized one, and while it succeeds warping there, then it imports 450386, 450387, 450388, 450389 from a reverted fork. What happens next is that it tries to import the legit 450390, but since it does not have the parent, it gets stuck
2025-11-24 09:06:06.767 WARN tokio-runtime-worker sync: [Container-2005] 💔 Error importing block 0x014c9158dac97aa1a94623e88f6ebb83b419821c05ac34a6873c7ff3c4fd60ef: block has an unknown parentAfter that, it gets stuck in idle, with the wrong best block
INFO tokio-runtime-worker substrate: [Container-2005] 💤 Idle (3 peers), best: #450389 (0xd936…7637), finalized #450385 (0xbcd7…9caf), ⬇ 108.9kiB/s ⬆ 6.6kiB/sThis happened while in stable2506, and it does not get fixed if restarted