[BUG] Infinite loop when using a negative value for job.schedule

[AlexRuiz7]

Describe the bug
The CM enters an infinite loop when a negative value is set for the job.schedule setting.

To Reproduce
Steps to reproduce the behavior:

• See Implement Command Manager settings on opensearch.yml #213 (comment)

Expected behavior

• No negative values are allowed for this setting.
• No infinite loop when using a custom value for this setting.
• The CM validates that this setting is under the expected working window.

Plugins
Command Manager.

Additional context
#213 (comment)

[f-galland]

Regarding the time values here, I believe they should keep the following relationship:

CLIENT_TIMEOUT < JOB_SCHEDULE
JOB_SCHEDULE == PIT_KEEP_ALIVE

The reasoning behind this is:

• CLIENT_TIMEOUT < JOB_SCHEDULE: The timeout is used for search actions. If the search timeout is longer than the period of the job, a concurrent search with the same query may be issued.
• JOB_SCHEDULE == PIT_KEEP_ALIVE: It is safer to keep the PIT alive throughout the job's lifecycle. Making it longer wouldn't make sense, since by the time a new job run is triggered, the data of the indices may have changed, so we need to issue a new PIT.