[
https://issues.apache.org/jira/browse/HDDS-17182?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Siyao Meng resolved HDDS-17182.
-------------------------------
Fix Version/s: 2.3.0
Resolution: Fixed
> Check key prefix and container ID when importing a container
> ------------------------------------------------------------
>
> Key: HDDS-17182
> URL: https://issues.apache.org/jira/browse/HDDS-17182
> Project: Apache Ozone
> Issue Type: Improvement
> Components: Ozone Datanode
> Reporter: Siyao Meng
> Assignee: Siyao Meng
> Priority: Major
> Labels: pull-request-available
> Fix For: 2.3.0
>
>
> When a datanode imports a container with schema V3,
> {{DatanodeStoreSchemaThreeImpl.loadKVContainerData}} writes every key of the
> RocksDB dump files in the archive into the volume's shared DB. Exporting
> ({{dumpKVContainerData}}) and cleanup ({{removeKVContainerData}}) are limited
> to the container's key prefix, but loading does not check it.
> {{ContainerImporter}} also takes the container ID from the {{.container}}
> file in the archive without comparing it with the requested container ID.
> This change:
> * Passes the container ID to {{loadKVContainerData}}, which fails on the
> first key outside that container's key prefix. The check runs before the
> batch is committed, so a failed load writes nothing.
> * Makes {{ContainerImporter}} reject an archive whose container ID differs
> from the requested one ({{INVALID_ARGUMENT}}).
> Compatibility:
> * No protocol, layout or configuration change.
> * Archives exported by datanodes are not affected, since they contain only
> keys of the exported container under its own ID.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
