GGUS: https://ggus.eu/tech/ticket_show.php?ticket=92758
Savannah: https://savannah.cern.ch/bugs/?86700
Savannah: https://savannah.cern.ch/bugs/?71100
CREAM running jobs whose proxy is expiring are killed. If a job is instead queued and its proxy expired, such killing is done only when the CREAM jw starts. CREAM instead should be able to kill such queued jobs as soon as the proxy expired. This is not a problem for jobs submitted through the WMS, since such killing is triggered by the WMS.
Approach for solving this issue: whenever a delegation proxy expires, all related jobs will be cancelled by CREAM. This simplify the logic and free the queue from jobs which have no way to be executed properly due to the expired proxy.