Requeued jobs do not retain their original age.
Dev ticket for https://gitlab.cern.ch/cta/operations/-/issues/593
When requeuing jobs, we set their age to zero again. This is unfair to the scheduler, which takes the oldest age of a job as a mount criteria. It means a mount may happen due to the oldest age criteria, fail to allocate disk space, requeue and have to wait for oldest age to rise again before being processed.