Using ClientRequestToken

[Chuukwudi]

AWS says that Job_Ids are stored for 7 days unless an output config bucket is provided.
Does this mean that when I spacify an output config bucket, I can still retrieve a job_id using the ClientRequestToken even after 10 days ?

Right now, I am trying to cache texract outputs for longer, I am not sure if I need an external db to to save and match idempotent ClientRequestToken to job_ids. in case textract really forgets after 7 days.

For the ClientRequestToken, I am using the md5 of the pdf_file.

Will texract guarantee returning me a job id after 7 days provided I spacify the output config.