INDY-1945: Implement NodeLeecherService #1120
Conversation
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
| self._state = self.State.SyncingAudit | ||
| self._leechers[AUDIT_LEDGER_ID].start(request_ledger_statuses) | ||
|
|
||
| def _on_ledger_leecher_stop(self, msg: LedgerCatchupComplete): |
There was a problem hiding this comment.
Is it used somewhere?
There was a problem hiding this comment.
It should be used, I missed adding message routing
| logger.warning("{} got unexpected catchup complete {} during idle state".format(self, msg)) | ||
|
|
||
| def _on_audit_synced(self, msg): | ||
| if msg.ledger_id != AUDIT_LEDGER_ID: |
There was a problem hiding this comment.
As of now, _current_ledger is used only when syncing other ledgers, and this is None when syncing Audit and Pool.
For a better consistency it would make sense to set it for every ledger.
It also would simplify a check at the beginning of _on_XXX_sycned methods.
There was a problem hiding this comment.
Actually I hope to get rid of _current_ledger once parallel catch-up is implemented, so I'm inclined to leave it as is for now.
| self._state = self.State.SyncingOthers | ||
| self._sync_next_ledger() | ||
|
|
||
| def _on_other_synced(self, msg): |
There was a problem hiding this comment.
Do we really need 3 on_synced methods?
I think we could create just one generic method (like the current _on_other_synced).
The only change is to set the state according to msg.ledger_id.
There was a problem hiding this comment.
I agree that this looks like some code duplication, however I wanted to make state machine as explicit as possible. I'll think about improving this.
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
Signed-off-by: Sergey Khoroshavin <sergey.khoroshavin@dsr-corporation.com>
| leecher = node.ledgerManager._node_leecher | ||
|
|
||
| # TODO: It does ignore designated ledger order, however it will obsolete after implemnting parallel catchup | ||
| ledger_ids = [ledger.id for ledger in node.ledgerManager.ledgerRegistry.values() |
There was a problem hiding this comment.
Why do we exclude Audit and Pool ledgers from checking?
There was a problem hiding this comment.
They are handled by separate states in NodeLeecherService, therefore they are not handled by get_next_ledger. Actually this function (and test) will be removed once parallel catch-up is implemented.
Signed-off-by: Sergey Khoroshavin sergey.khoroshavin@dsr-corporation.com