feat(resharding): implemented error handling and test for it #10179
Conversation
wacban
force-pushed
the
waclaw-resharding-nayduck-error
branch
from
df1715f
to
ea77618
Compare
| }) | ||
|
|
||
| let state_snapshot = tries.get_state_snapshot(prev_prev_hash); | ||
| if let Err(err) = state_snapshot { |
There was a problem hiding this comment.
Was the old code too convoluted to understand?
There was a problem hiding this comment.
not really, I just wanted to add that debug line in the middle of it
| @@ -1041,7 +1041,7 @@ impl StateSync { | |||
| shard_id, | |||
| state_split_scheduler, | |||
| )?; | |||
| tracing::debug!(target: "sync", %shard_id, %sync_hash, ?me, "State sync split scheduled"); | |||
| tracing::debug!(target: "sync", %shard_id, %sync_hash, ?me, "resharding scheduled"); | |||
There was a problem hiding this comment.
aah, how did we miss this in the past?
tools/database/src/corrupt.rs
Outdated
| let flat_storage_manager = FlatStorageManager::new(store.clone()); | ||
|
|
||
| let mut store_update = store.store_update(); | ||
| // TODO(resharding) there must be a better way to get the shard uids |
There was a problem hiding this comment.
This highly depends on what shard version we are trying to corrupt right? Could we pass in the shard_version as input here maybe and use that to get the shard_uids?
There was a problem hiding this comment.
Not a bad workaround at all, though I prefer to keep the cmdline as user friendly as possible.
It's not a top priority right now so I'll leave that in and we can clean up after all else is done.
|
https://nayduck.near.org/#/run/3273 |
https://nayduck.near.org/#/run/3270