Documentation ¶
Index ¶
- Constants
- Variables
- func RunHostProvisionFailTriggers(h *host.Host) error
- func RunLastRevisionNotFoundTrigger(proj *model.ProjectRef, v *version.Version) error
- func RunSpawnWarningTriggers(host *host.Host) error
- func RunTaskFailureTriggers(taskId string) error
- type AlertContext
- type Deliverer
- type EmailDeliverer
- type FirstFailureInTaskType
- type FirstFailureInVariant
- type FirstFailureInVersion
- type LastRevisionNotFound
- type ProvisionFailed
- type QueueProcessor
- type SMTPSettings
- type SlowProvisionWarning
- type SpawnFailure
- type SpawnTwelveHourWarning
- type SpawnTwoHourWarning
- type TaskFailTransition
- type TaskFailed
- type Trigger
Constants ¶
const ( EmailProvider = "email" JiraProvider = "jira" SlackProvider = "slack" RunnerName = "alerter" )
const DescriptionTemplateString = `` /* 451-byte string literal not displayed */
DescriptionTemplateString defines the content of the alert ticket.
const (
EmailSubjectPrologue = "[Evergreen]"
)
Variables ¶
var ( // AvailableTaskTriggers is a list of all the supported task triggers, which is used by the UI // package to generate a control panel for configuring how to react to these triggers. AvailableTaskFailTriggers = []Trigger{TaskFailed{}, FirstFailureInVersion{}, FirstFailureInVariant{}, FirstFailureInTaskType{}, TaskFailTransition{}, } AvailableProjectTriggers = []Trigger{ LastRevisionNotFound{}, } SpawnWarningTriggers = []Trigger{SpawnTwoHourWarning{}, SpawnTwelveHourWarning{}} )
var DescriptionTemplate = template.Must(template.New("Desc").Parse(DescriptionTemplateString))
DescriptionTemplate is filled to create a JIRA alert ticket. Panics at start if invalid.
Functions ¶
func RunLastRevisionNotFoundTrigger ¶
func RunLastRevisionNotFoundTrigger(proj *model.ProjectRef, v *version.Version) error
func RunSpawnWarningTriggers ¶
func RunTaskFailureTriggers ¶
RunTaskTriggers queues alerts for any active triggers on the tasks's state change.
Types ¶
type AlertContext ¶
type AlertContext struct { AlertRequest *alert.AlertRequest ProjectRef *model.ProjectRef Task *task.Task Build *build.Build Version *version.Version Patch *patch.Patch Host *host.Host FailedTests []task.TestResult Settings *evergreen.Settings }
AlertContext is the set of full documents in the DB that are associated with the values found in a given AlertRequest
type Deliverer ¶
type Deliverer interface {
Deliver(AlertContext, model.AlertConfig) error
}
Deliverer is an interface which handles the actual delivery of an alert. (e.g. sending an e-mail, posting to flowdock, etc)
type EmailDeliverer ¶
type EmailDeliverer struct { SMTPSettings // contains filtered or unexported fields }
EmailDeliverer is an implementation of Deliverer that sends notifications to an SMTP server
func (*EmailDeliverer) Deliver ¶
func (es *EmailDeliverer) Deliver(alertCtx AlertContext, alertConf model.AlertConfig) error
type FirstFailureInTaskType ¶
type FirstFailureInTaskType struct{}
FirstFailureInVersion is a trigger that queues an alert whenever a task fails for the first time for a task of a given name within a version. For example: "compile" fails on linux-64; ShouldExecute returns true "compile" fails on windows; ShouldExecute returns false because one was already sent for compile. "unit-tests" fails on windows; ShouldExecute returns true because nothing was sent yet for unit-tests. "unit-tests" fails on linux-64; ShouldExecute returns false
func (FirstFailureInTaskType) CreateAlertRecord ¶
func (trig FirstFailureInTaskType) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (FirstFailureInTaskType) Display ¶
func (trig FirstFailureInTaskType) Display() string
func (FirstFailureInTaskType) Id ¶
func (trig FirstFailureInTaskType) Id() string
func (FirstFailureInTaskType) ShouldExecute ¶
func (trig FirstFailureInTaskType) ShouldExecute(ctx triggerContext) (bool, error)
type FirstFailureInVariant ¶
type FirstFailureInVariant struct{}
FirstFailureInVersion is a trigger that queues an alert whenever a task fails for the first time within a variant. After one failure has triggered an alert for this event, subsequent failures will not trigger additional alerts.
func (FirstFailureInVariant) CreateAlertRecord ¶
func (trig FirstFailureInVariant) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (FirstFailureInVariant) Display ¶
func (trig FirstFailureInVariant) Display() string
func (FirstFailureInVariant) Id ¶
func (trig FirstFailureInVariant) Id() string
func (FirstFailureInVariant) ShouldExecute ¶
func (trig FirstFailureInVariant) ShouldExecute(ctx triggerContext) (bool, error)
type FirstFailureInVersion ¶
type FirstFailureInVersion struct{}
FirstFailureInVersion is a trigger that queues an alert whenever a task fails for the first time within a version. After one failure has triggered an alert for this event, subsequent failures will not trigger additional alerts.
func (FirstFailureInVersion) CreateAlertRecord ¶
func (trig FirstFailureInVersion) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (FirstFailureInVersion) Display ¶
func (trig FirstFailureInVersion) Display() string
func (FirstFailureInVersion) Id ¶
func (trig FirstFailureInVersion) Id() string
func (FirstFailureInVersion) ShouldExecute ¶
func (trig FirstFailureInVersion) ShouldExecute(ctx triggerContext) (bool, error)
type LastRevisionNotFound ¶
type LastRevisionNotFound struct{}
func (LastRevisionNotFound) CreateAlertRecord ¶
func (lrnf LastRevisionNotFound) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (LastRevisionNotFound) Display ¶
func (lrnf LastRevisionNotFound) Display() string
func (LastRevisionNotFound) Id ¶
func (lrnf LastRevisionNotFound) Id() string
func (LastRevisionNotFound) ShouldExecute ¶
func (lrnf LastRevisionNotFound) ShouldExecute(ctx triggerContext) (bool, error)
type ProvisionFailed ¶
type ProvisionFailed struct{}
func (ProvisionFailed) CreateAlertRecord ¶
func (pf ProvisionFailed) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (ProvisionFailed) Display ¶
func (pf ProvisionFailed) Display() string
func (ProvisionFailed) Id ¶
func (pf ProvisionFailed) Id() string
func (*ProvisionFailed) ShouldExecute ¶
func (pf *ProvisionFailed) ShouldExecute(ctx triggerContext) (bool, error)
type QueueProcessor ¶
type QueueProcessor struct {
// contains filtered or unexported fields
}
QueueProcessor handles looping over any unprocessed alerts in the queue and delivers them.
This runner is used for build and enqueue failure notifications
func (*QueueProcessor) Deliver ¶
func (qp *QueueProcessor) Deliver(req *alert.AlertRequest, ctx *AlertContext) error
func (*QueueProcessor) Name ¶
func (qp *QueueProcessor) Name() string
type SMTPSettings ¶
type SlowProvisionWarning ¶
type SlowProvisionWarning struct{}
func (SlowProvisionWarning) CreateAlertRecord ¶
func (sthw SlowProvisionWarning) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (SlowProvisionWarning) Display ¶
func (sthw SlowProvisionWarning) Display() string
func (SlowProvisionWarning) Id ¶
func (sthw SlowProvisionWarning) Id() string
func (SlowProvisionWarning) ShouldExecute ¶
func (sthw SlowProvisionWarning) ShouldExecute(ctx triggerContext) (bool, error)
type SpawnFailure ¶
type SpawnFailure struct{}
func (SpawnFailure) CreateAlertRecord ¶
func (sf SpawnFailure) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (SpawnFailure) Display ¶
func (sf SpawnFailure) Display() string
func (SpawnFailure) Id ¶
func (sf SpawnFailure) Id() string
func (SpawnFailure) ShouldExecute ¶
func (sf SpawnFailure) ShouldExecute(ctx triggerContext) (bool, error)
type SpawnTwelveHourWarning ¶
type SpawnTwelveHourWarning struct{}
func (SpawnTwelveHourWarning) CreateAlertRecord ¶
func (sthw SpawnTwelveHourWarning) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (SpawnTwelveHourWarning) Display ¶
func (sthw SpawnTwelveHourWarning) Display() string
func (SpawnTwelveHourWarning) Id ¶
func (sthw SpawnTwelveHourWarning) Id() string
func (SpawnTwelveHourWarning) ShouldExecute ¶
func (sthw SpawnTwelveHourWarning) ShouldExecute(ctx triggerContext) (bool, error)
type SpawnTwoHourWarning ¶
type SpawnTwoHourWarning struct{}
func (SpawnTwoHourWarning) CreateAlertRecord ¶
func (sthw SpawnTwoHourWarning) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (SpawnTwoHourWarning) Display ¶
func (sthw SpawnTwoHourWarning) Display() string
func (SpawnTwoHourWarning) Id ¶
func (sthw SpawnTwoHourWarning) Id() string
func (SpawnTwoHourWarning) ShouldExecute ¶
func (sthw SpawnTwoHourWarning) ShouldExecute(ctx triggerContext) (bool, error)
type TaskFailTransition ¶
type TaskFailTransition struct{}
TaskFailTransition is a trigger that queues an alert iff the following conditions are met: 1) A task fails and the previous completion of this task on the same variant was passing or the task has never run before 2) The most recent alert for this trigger, if existing, was stored when the 'last passing task' at the time was older than the 'last passing task' for the newly failed task. 3) The previous run was a failure, and there has been Multipler*Batchtime time since the previous alert was sent.
func (TaskFailTransition) CreateAlertRecord ¶
func (trig TaskFailTransition) CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord
func (TaskFailTransition) Display ¶
func (trig TaskFailTransition) Display() string
func (TaskFailTransition) Id ¶
func (trig TaskFailTransition) Id() string
func (TaskFailTransition) ShouldExecute ¶
func (trig TaskFailTransition) ShouldExecute(ctx triggerContext) (bool, error)
type TaskFailed ¶
type TaskFailed struct{}
TaskFailed is a trigger that queues an alert whenever a task fails, regardless of any alerts generated by previous runs of the task or other tasks within the version/variant/task type.
func (TaskFailed) CreateAlertRecord ¶
func (trig TaskFailed) CreateAlertRecord(_ triggerContext) *alertrecord.AlertRecord
func (TaskFailed) Display ¶
func (trig TaskFailed) Display() string
func (TaskFailed) Id ¶
func (tf TaskFailed) Id() string
func (TaskFailed) ShouldExecute ¶
func (trig TaskFailed) ShouldExecute(ctx triggerContext) (bool, error)
type Trigger ¶
type Trigger interface { // Id is a string that identifies this trigger logic. It's used to store in the database for // alerts that are queued to identify the reason the alert was generated. Id() string // Display is a human-readable description of the trigger logic. It may be used in log messages // for debugging or for presentation in a UI control panel. Display() string // ShouldExecute returns a boolean indicating if this trigger's logic has matched and should // result in an alert being queued. This logic may involve querying for bookkeeping // generated by CreateBookkeepingRecord in previous executions. ShouldExecute(ctx triggerContext) (bool, error) // CreateAlertRecord returns an instance of AlertRecord which, after insertion, should allow // this trigger to detect previous executions of itself in its ShouldExecute function. CreateAlertRecord(ctx triggerContext) *alertrecord.AlertRecord }
Trigger is a rule that determines if an alert should be queued for some event.