[
https://issues.apache.org/jira/browse/IGNITE-22624?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Alexander Lapin updated IGNITE-22624:
-------------------------------------
Description: There's a race between adding a table processor to replicas
and processing a request targeted to that table. Thus we may catch NPE on
replicas.get(tableId). In order to fix that we may use schema sync mechanism
for all effectively TableAware requests. (was: The same mechanism must be
implemented for the table replica listeners.)
> Await for the zone replica future ready on addTableReplicaListener
> ------------------------------------------------------------------
>
> Key: IGNITE-22624
> URL: https://issues.apache.org/jira/browse/IGNITE-22624
> Project: Ignite
> Issue Type: Improvement
> Reporter: Kirill Gusakov
> Assignee: Alexander Lapin
> Priority: Major
> Labels: ignite-3
>
> There's a race between adding a table processor to replicas and processing a
> request targeted to that table. Thus we may catch NPE on
> replicas.get(tableId). In order to fix that we may use schema sync mechanism
> for all effectively TableAware requests.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)