HDDS-15150. Container scanner should not mark container as UNHEALTHY when FD exhausted

[sarvekshayr]

What changes were proposed in this pull request?

Fixed a bug where background container scanners marked containers as UNHEALTHY due to resource issues rather than actual data corruption. Specifically, when the system encountered a FileNotFoundException or FileSystemException caused by file-descriptor exhaustion ("Too many open files"), the scanners incorrectly flagged these as corruption events.

The logic has been updated to explicitly catch these resource-related exceptions, ensuring that containers remain in their current state when the scanner cannot perform its check due to system limits.

ContainerMetadataScanner

2026-04-20 22:01:43,978 ERROR [ContainerMetadataScanner]-org.apache.hadoop.ozone.container.ozoneimpl.BackgroundContainerMetadataScanner: Corruption detected in container [3980819]. Marking it UNHEALTHY.
java.io.FileNotFoundException: /data6/hadoop-ozone/datanode/data/hdds/CID-637fe7c5-f40b-4e49-98b3-52154bd669e2/current/containerDir95/3980819/metadata/3980819.container (Too many open files)

ContainerDataScanner

2026-04-20 22:01:43,982 ERROR [ContainerDataScanner(/data12/hadoop-ozone/datanode/data/hdds)]-org.apache.hadoop.ozone.container.ozoneimpl.BackgroundContainerDataScanner: Corruption detected in container [16326340]. Marking it UNHEALTHY.
java.nio.file.FileSystemException: /data12/hadoop-ozone/datanode/data/hdds/CID-637fe7c5-f40b-4e49-98b3-52154bd669e2/current/containerDir143/16326340/chunks/115816904944438982.block: Too many open files

What is the link to the Apache JIRA

HDDS-15150

How was this patch tested?

Added unit tests in TestBackgroundContainerDataScanner and TestBackgroundContainerMetadataScanner.
Verified that with fix, containers are not incorrectly marked as UNHEALTHY.