Documentation ¶
Overview ¶
Package summarizer provides a method to read state protos defined in a config an output summary protos.
Index ¶
- Constants
- func CalculateHealthiness(grid *statepb.Grid, startTime int, endTime int, tab string) *summarypb.HealthinessInfo
- func CalculateTrend(currentHealthiness, previousHealthiness *summarypb.HealthinessInfo)
- func Update(ctx context.Context, client gcs.ConditionalClient, mets *Metrics, ...) error
- type Metrics
Constants ¶
const (
// DefaultInterval is the default number of days of analysis
DefaultInterval = 7
)
Variables ¶
This section is empty.
Functions ¶
func CalculateHealthiness ¶ added in v0.0.13
func CalculateHealthiness(grid *statepb.Grid, startTime int, endTime int, tab string) *summarypb.HealthinessInfo
CalculateHealthiness extracts the test run data from each row (which represents a test) of the Grid and then analyzes it with an implementation of flakinessAnalyzer, which has implementations in the subdir naive and can be injected as needed.
func CalculateTrend ¶ added in v0.0.14
func CalculateTrend(currentHealthiness, previousHealthiness *summarypb.HealthinessInfo)
CalculateTrend populates the ChangeFromLastInterval fields of each TestInfo by comparing the current flakiness to the flakiness calculated for the last interval. Interval length is a config value that is 7 days by default. The Trend enum defaults to UNKNOWN, so there is no need to explicitly assign UNKNOWN when a test appears in currentHealthiness but not in previousHealthiness.
func Update ¶
func Update(ctx context.Context, client gcs.ConditionalClient, mets *Metrics, configPath gcs.Path, concurrency int, dashboard, gridPathPrefix, summaryPathPrefix string, confirm bool) error
Update summary protos by reading the state protos defined in the config.
Will use concurrency go routines to update dashboards in parallel. Setting dashboard will limit update to this dashboard. Will write summary proto when confirm is set.
Types ¶
Directories ¶
Path | Synopsis |
---|---|
Package analyzers represents ways to analyze healthiness and flakiness of tests Package analyzers represents ways to analyze healthiness and flakiness of tests
|
Package analyzers represents ways to analyze healthiness and flakiness of tests Package analyzers represents ways to analyze healthiness and flakiness of tests |