{"name":"RMAN Archived-Log Backup Failures (24h)","description":"Counts failed archived-log backup jobs recorded in RMAN's job history within the last 24 hours. Archived-log backups typically run far more often than full backups, so a single missed run is usually low-impact on its own, but repeated failures point to a problem with the backup destination, such as a full Fast Recovery Area, that will eventually block redo generation and hang the instance. Consider pairing this metric with a Fast Recovery Area space-usage metric so the two alerts corroborate each other.","tsql":"-- Returns the count of failed RMAN archived-log backup jobs in the last 24 hours\r\n-- RMAN backup jobs operate at the whole-CDB level, reported at instance level (con_id 0)\r\nSELECT 0 AS con_id,\r\n       COUNT(*) AS metric_value,\r\n       LISTAGG('started ' || TO_CHAR(start_time, 'DD-MON HH24:MI') || ' ' || status,\r\n               '; ' ON OVERFLOW TRUNCATE)\r\n           WITHIN GROUP (ORDER BY start_time DESC) AS alert_detail_text\r\n  FROM v$rman_backup_job_details\r\n WHERE input_type = 'ARCHIVELOG'\r\n   AND status = 'FAILED'\r\n   AND start_time > SYSDATE - 1","instances":true,"databases":{"mode":1},"frequency":1800,"frequency_comment":"seconds (collected every 30 minutes)","rateofchange":false,"metricenabled":true,"alertname":"RMAN archived-log backup failed","alertdescription":"One or more RMAN archived-log backup jobs failed within the last 24 hours, breaking the point-in-time recovery chain. Repeated failures point to a backup destination problem, such as a filling Fast Recovery Area, that will eventually block redo generation and hang the instance.","aboveorbelow":"above","_thresholds_high":{"selected":true,"value":0},"collections":1,"alertenabled":true,"_thresholds_medium":{"selected":false},"_thresholds_low":{"selected":false},"targetentitytype":2}