ezgh obs alerts list [--state firing|pending|noData|silenced]... [--severity s]... [--label name=value]... [--owner rule:<rule>]... [--grouped] [flags]List the project’s alert instances: each label set of a rule (or an SLO’s tier) that is pending, firing or without data, and those resolved in the last 15 minutes. Firing ones come first, then those without data, pending and silenced ones, each newest first. SILENCED says whether an active silence mutes its notifications (evaluation goes on regardless); --state silenced keeps those, and the other states leave them out. An SLO’s instances show as slo:<name> (tier).
--state, --severity (the severity label), --label (name=value), --owner (rule:<rule> or slo:<id>) and --search (part of the rule’s or SLO’s name) narrow the list; each but --search can be given more than once. The list is paged (--limit, default 50; --all reads every page).
--grouped lists one group per rule (or SLO tier) instead, worst first: its counts by state, the labels all its instances share, and its first 10 instances, with only their other labels. Above the groups are the project’s counts by state (with every filter but --state) and the rules in rule error. --limit is then groups per page (1 to 50, default 20).
The labels every instance carries (alertname, ezgh_rule) are left out of the table; -o json has them.
Part of ezgh obs alerts.
Examples
ezgh obs alerts list
ezgh obs alerts list --state firing --severity critical --label service=api
ezgh obs alerts list --owner rule:"High API latency" --all -o json
ezgh obs alerts list --groupedFlags
| Flag | Type | Default | Description |
|---|---|---|---|
--all |
boolean | follow nextCursor through every page | |
--cursor |
string | start from this page (a previous page’s nextCursor) | |
--grouped |
boolean | list one group per rule (or SLO tier), with its first 10 instances | |
--label |
string, repeatable | only instances with this label: name=value (repeatable) | |
--limit |
int | 50 |
items per page, 1 to 100 |
--owner |
string, repeatable | only these rules’ or SLOs’ instances: rule:<rule> (ID, slug or name) or slo:<id> (repeatable) | |
--rule |
string | only this rule’s instances (ID, slug or name); the same as --owner rule:<rule> |
|
--search |
string | only rules and SLOs whose name contains this (ignoring case) | |
--severity |
strings, comma-separated or repeated | only instances whose severity label is one of these (repeat, or comma-separate) | |
--state |
strings, comma-separated or repeated | only instances in these states: firing, pending, noData, silenced or (without --grouped) resolved (repeat, or comma-separate) |
This command also takes the global flags.