Skip to content

Commit

Permalink
critical alerts: waiting jobs in queue (#732)
Browse files Browse the repository at this point in the history
  • Loading branch information
alpetric authored Oct 10, 2024
1 parent fe51fb1 commit 14c735a
Show file tree
Hide file tree
Showing 3 changed files with 5 additions and 0 deletions.
1 change: 1 addition & 0 deletions docs/core_concepts/37_critical_alert_channels/index.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -14,6 +14,7 @@ Critical alerts are generated under the following conditions:
- [License key](../../misc/7_plans_details/index.mdx#using-the-license-key-self-host) does not renew.
- [Workspace error handler](../10_error_handling/index.mdx#workspace-error-handler) fails.
- Number of running workers in a group falls below a specified threshold (has to be configured in the [worker group](../9_worker_groups/index.mdx) config).
- Number of [jobs waiting in queue](../9_worker_groups/index.mdx#queue-metric-alerts) is above a threshold for more than a specified amount of time.

You just need to [configure SMTP](../../advanced/1_self_host/index.mdx#set-up-smtp-from-the-ui) and setup a critical alert channel (aka email address) in the [instance settings](../../advanced/18_instance_settings/index.mdx#smtp) or connect your instance to Slack and fill in a channel name.

Expand Down
4 changes: 4 additions & 0 deletions docs/core_concepts/9_worker_groups/index.mdx
Original file line number Diff line number Diff line change
Expand Up @@ -321,6 +321,10 @@ Only tags for jobs that have been delayed by more than 3 seconds in the last 14
![Queue Metrics](./queue_metrics.png 'Queue Metrics')
### Queue Metric Alerts
Enterprise Edition users can set up [Critical Alerts](../37_critical_alert_channels/index.mdx) on the [Queue Metrics page](#queue-metrics), and be notified when the number of delayed jobs in a queue is above a certain threshold for more than a configured amount of time. The "cooldown" parameter determines the minimum duration between two consecutive alerts if the number of waiting jobs are fluctuating around the configured threshold.
![Queue Metrics](./queue_metrics_alert.png 'Queue Metrics Alert')
## Workers and vCPUs
Even though Windmill's architecture relies on workers, Windmill's [pricing](/pricing) is vCPU-based. For example, 4 workers with 0.25 vCPU each is 1 vCPU. 1 worker with 4 vCPU would count as 4 vCPU. But if it only runs for 1h every day, you would divide that by 24. Each vCPU can run up to ~26M jobs per month.
Expand Down
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.

0 comments on commit 14c735a

Please sign in to comment.