clp-package: Fix several search job cancellation issues:

[wraymo]

References

Description

This PR addresses several cancellation issues:

• Previously, when all search tasks completed, sending a cancellation request changed the status to CANCELLING. Since the job was already removed from active_jobs, it remained in an intermediate state. This PR ensures that a cancellation request is only sent if the job status is PENDING or RUNNING. Besides, for completed jobs in RUNNING status, even if the cancellation request was sent, the job status is updated to either SUCCEEDED or FAILED based on the actual completion.
• The wrong type was passed when handling job cancellation, so that the code never reached cancel_job_except_reducer. This PR fixes this issue.

Validation performed

• Started the package and compressed the hive-24hrs dataset.
• In the command line, executed sbin/search.sh "org.apache.hadoop.metrics2.impl.MetricsConfig: loaded properties from hadoop-metrics2.properties" and checked search_jobs table in the database.
• Ran the same query, and cancelled it before it finished. Checked search_jobs table in the database.

[kirkrodrigues]

Nice catches! A few minor suggestions.

[kirkrodrigues]

Suggested change

	WHERE ${SEARCH_JOBS_TABLE_COLUMN_NAMES.ID} = ?
	AND ${SEARCH_JOBS_TABLE_COLUMN_NAMES.STATUS}
	WHERE ${SEARCH_JOBS_TABLE_COLUMN_NAMES.ID} = ?
	AND ${SEARCH_JOBS_TABLE_COLUMN_NAMES.STATUS}

[kirkrodrigues]

For the PR title/commit description, how about:

clp-package: Fix several search job cancellation issues:
- Only allow cancelling pending or running jobs from the webui.
- Set a job's status completion status regardless of whether it was concurrently cancelled.
- Cast search job ID appropriately before indexing the in-memory active jobs dict.