Monitor Git provider rate limits
The Rate limit consumption section of the Observability tab provides visibility into Git provider API rate-limit usage across your Git connectors. Use this view to monitor API consumption, identify connectors approaching their rate limits, and take corrective action before rate-limit exhaustion affects Git operations.
Git providers enforce API rate limits on requests made within a given time window. This view helps you track API usage by connector so you can avoid disruptions to Git operations caused by rate-limit exhaustion.
If your Git connector authenticates through a Harness Delegate, the delegate must be running version 896xx or later to use the rate limit consumption feature.
What will you learn in this topic?
This topic explains how to monitor Git provider API rate limits. It covers:
- How to view rate-limit consumption by connector.
- How to interpret rate-limit metrics.
- How to view historical API usage.
- How to filter and customize the rate-limit view.
View rate-limit consumption
To access this view, go to the Webhooks page, select the Observability tab, and expand the Rate limit consumption section.
The rate-limit metrics displayed in this view include only Git provider API calls made by Harness using the configured Git connector credentials. API requests made outside of Harness (for example, by users, scripts, CI/CD systems, or other applications using the same credentials) are not tracked and are not reflected in the reported consumption.
Rate-limit data is retained for the last one day only. If no Git API requests have been made by Harness using a connector during this period, the stored data is reset, and rate-limit data is unavailable for that connector. In this case, the UI displays the default (-) or null values.
To scope and customize the view:
- Use the Search by connector field and the Provider filter to scope the list.
- The Provider filter supports All providers, GitHub, GitLab, Bitbucket, Bitbucket Server, and Azure Repos.
- Select the refresh icon to reload the metrics, and use the Columns selector to show or hide columns (select Reset to restore the defaults). The active filters are reflected in the page URL, so you can bookmark or share a specific view.
Each row shows a connector's consumption for a rate-limit bucket:
| Column | Description |
|---|---|
| Connector | The connector consuming the Git provider API. |
| Provider | The Git provider for the connector. |
| Bucket | The rate-limit bucket, for example core. Harness shows a bucket for each rate-limit category the provider supports. For example, GitHub exposes buckets such as core and search, while providers like Bitbucket do not expose named buckets, so their consumption is shown under a single default bucket. |
| Current consumption | A progress bar and the percentage of the limit currently used. |
| Used / Limit | Requests used against the maximum allowed, for example 29 / 5000. |
| Remaining | The number of requests still available in the current window. |
| Resets in | When the current rate-limit window resets. |
| Throttled requests | The number of requests blocked because the limit was reached. |
| Last updated | When Harness last refreshed the metrics for the connector. |
View historical API usage
To see historical usage:
- Expand a connector row or select View details from the ⋮ menu to display the consumption chart.
- To retrieve the latest metrics, select Refresh from the ⋮ menu in the expanded view.
- Use the Window selector to change the time range:
- Last 1h
- Last 3h
- Last 6h
- Last 12h
- Last 24h
- Hover over a data point to view the bucket, consumption value, and timestamp.
If Harness has no data for the selected connector and time window, the chart displays No timeline data.
This view is behind the feature flag PIPE_GITX_RATE_LIMITS_API. Contact Harness Support to enable the feature.
Next steps
- Monitor repository sync health: Monitor webhook coverage and synchronization status.
- Git rate limit best practices: Reduce Git provider API consumption and avoid rate-limit exhaustion.