This does not account for a failed status. How about instead using an enum:

enum SearchStatus {
    Pending = 'pending',
    Completed = 'completed',
    Failed = 'failed'
}

@Entity()
export class SearchStat {
  ...
  @Column({ type: 'text' })
  status: SearchStatus;
}

I did not consider the failed status, as in the queue used max attempts and exponential backoff delay to make sure the failed job retries continue with exponential backoff. But there can be cases, the job exhausts all the retries and unable to process the search. Should have considered this scenerio

In that case, when the queued job exhausts all its attempts and is unable to process the search operation, I guess in that scenario I need to add a cron job scheduler to pick all failed jobs and process them again?

In a production application, taking care of the failed jobs could take many forms (retrying them, or sending a report/error), but it is outside of the scope of this exercise.

As for the exponential backoff, still, the status could be marked as failed in the mean time to differentiate it from other jobs 💡