---
url: /changelog/v9.6.x/prometheus-migration.md
---

# Prometheus Metrics Migration

With the release of Teamscale 9.6, we reworked some of the metrics which can be exposed via the [Prometheus Metrics Service](/howto/monitoring-teamscale/#api-endpoint-for-prometheus).

* teamscale\_voting\_time → voting\_time\_seconds
  * Unit changed: ms → seconds
* cpu\_logical\_cores → cpu\_core
* cpu\_load → cpu\_load\_ratio
* ram\_installed → ram{type="total|used"}
* ram\_usage → removed, calculate from ram\_installed instead.
* teamscale\_precommit\_latency\_seconds → precommit\_latency\_seconds
* teamscale\_precommit\_polling\_total → precommit\_polling
* teamscale\_precommit\_open\_tasks\_total → precommit\_open\_tasks
* teamscale\_precommit\_content\_total → precommit\_content
* teamscale\_precommit\_unique\_users\_last\_five\_minutes\_total → precommit\_unique\_users\_last\_five\_minutes
* store\_open\_total → store\_open
* store\_operation\_total → store\_operation
* store\_operation\_key\_total → store\_operation\_key
* workers\_count → workers
* workers\_load → workers\_load\_ratio
* worker\_runtime → worker\_runtime\_seconds
  * Unit changed: ms → seconds
* license\_expires → license\_days\_left
* teamscale\_commit\_analysis\_time → commit\_analysis\_duration\_seconds
  * Unit changed: ms → seconds
* string\_abbreviator\_cache\_hits, string\_abbreviator\_cache\_misses → string\_abbreviator\_cache{type="hit|miss"}
* string\_abbreviator\_cache\_hit\_percent → removed, calculate from string\_abbreviator\_cache instead.
* string\_abbreviator\_cache\_estimated\_size → string\_abbreviator\_cache\_size\_bytes{type="estimated|max"}
