Closed anjackson closed 2 years ago
Also, may as well avoid this one:
Labels
alertname = host_free_disk_space_jisc03
device = /dev/mapper/vg_local-lv_root
fstype = ext4
instance = jisc03:9100
job = solr
mountpoint = /
severity = severe
Done.
Crawls are now launched with Airflow, so
alertname=crawl_launcher_has_not_run
alerts are no longer necessary. Airflow itself manages jobs and alerts on failures, so no replacement alert is needed in Prometheus itself.