The compute cluster will be down for routine maintenance from October 13 until October 14. Most jobs will be queued until the cluster is back up.