Skip to content

Do not fire ContainerMemoryUsage alert for unlabelled metrics

The prometheus query we use for ContainerMemoryUsage alert

(sum(container_memory_working_set_bytes) BY (container) / sum(container_spec_memory_limit_bytes > 0) BY (container) * 100) > 80

regularly leads to alerts that do not have the container label set. It seems that those metrics are redundant and/or aggregate versions of the ones that do have the container label. Therefore we should adapt the query to leave out those unlabelled metrics.

To upload designs, you'll need to enable LFS and have an admin enable hashed storage. More information