fix: add Evicted event

[ZhangSetSail]

No description provided.

[Copilot]

Pull request overview

This PR enhances pod eviction event handling by extending the monitoring window and adding more specific eviction reason detection. The changes enable the system to detect and report eviction events throughout a pod's entire lifecycle rather than just within the first 30 minutes.

• Extended monitoring window from 30 minutes to indefinite to catch runtime evictions
• Added granular eviction reason detection for PID exhaustion, inode exhaustion, and ephemeral storage issues
• Enhanced error messages with more context and user guidance

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
worker/master/podevent/podevent.go	Removed 30-minute monitoring upper bound, extended cache cleanup to 24 hours, and enhanced eviction detection with specific resource types
util/language.go	Added Chinese translations for new eviction event types (PID, inode, disk pressure, generic resource pressure)

[Copilot]

The translation key used for memory eviction is inconsistent with the eviction context. For memory evictions, the message uses "Deployment failed: container out of memory killed" which is actually meant for OOMKilled containers (line 656), not pod evictions. This should use a dedicated translation key like "Deployment failed: pod evicted due to memory pressure" to maintain consistency with the other eviction-specific messages.

Suggested change

	msg = fmt.Sprintf("%s: %s", util.Translation("Deployment failed: container out of memory killed"), pod.Status.Message)
	msg = fmt.Sprintf("%s: %s", util.Translation("Deployment failed: pod evicted due to memory pressure"), pod.Status.Message)