Cron or a systemd timer: what happens when the job fails
Both run a job on a schedule. They differ in what happens when the machine was off, when the last run is still going, where the output ends up, and who hears about a failure.
Topic · 7 stories
7 stories tagged scheduling on AI Tamer, newest first.
Both run a job on a schedule. They differ in what happens when the machine was off, when the last run is still going, where the output ends up, and who hears about a failure.
A replica count alone does not spread Pods across failure zones. Kubernetes topology spread constraints let you define the placement you need and the tradeoff when a zone is unavailable.
Use a PostgreSQL exclusion constraint to keep active reservations for the same GPU from overlapping. Range bounds and cancellation rules make the schedule explicit.
Kubernetes can give urgent Pods earlier scheduling attempts without removing running work. Non-preempting priority still depends on suitable capacity.
CPU requests guide Pod placement. CPU limits cap a running container's CPU time.
Device plugins advertise GPUs as countable resources. That changes how Pods request them and how many workloads a node can schedule.
A fixed timestamp and a recurring local time keep different clocks steady. Choosing the right one prevents a daylight-saving surprise.