grafana/pkg/services/ngalert/api/api_ruler.go

284 lines
9.9 KiB
Go
Raw Normal View History

package api
import (
"errors"
"fmt"
"net/http"
"time"
"github.com/grafana/grafana/pkg/services/datasources"
"github.com/grafana/grafana/pkg/services/ngalert/state"
"github.com/grafana/grafana/pkg/services/ngalert/store"
coreapi "github.com/grafana/grafana/pkg/api"
"github.com/grafana/grafana/pkg/api/response"
"github.com/grafana/grafana/pkg/infra/log"
"github.com/grafana/grafana/pkg/models"
Inhouse alerting api (#33129) * init * autogens AM route * POST dashboards/db spec * POST alert-notifications spec * fix description * re inits vendor, updates grafana to master * go mod updates * alerting routes * renames to receivers * prometheus endpoints * align config endpoint with cortex, include templates * Change grafana receiver type * Update receivers.go * rename struct to stop swagger thrashing * add rules API * index html * standalone swagger ui html page * Update README.md * Expose GrafanaManagedAlert properties * Some fixes - /api/v1/rules/{Namespace} should return a map - update ExtendedUpsertAlertDefinitionCommand properties * am alerts routes * rename prom swagger section for clarity, remove example endpoints * Add missing json and yaml tags * folder perms * make folders POST again * fix grafana receiver type * rename fodler->namespace for perms * make ruler json again * PR fixes * silences * fix Ok -> Ack * Add id to POST /api/v1/silences (#9) Signed-off-by: Ganesh Vernekar <cs15btech11018@iith.ac.in> * Add POST /api/v1/alerts (#10) Signed-off-by: Ganesh Vernekar <cs15btech11018@iith.ac.in> * fix silences * Add testing endpoints * removes grpc replace directives * [wip] starts validation * pkg cleanup * go mod tidy * ignores vendor dir * Change response type for Cortex/Loki alerts * receiver unmarshaling tests * ability to split routes between AM & Grafana * api marshaling & validation * begins work on routing lib * [hack] ignores embedded field in generation * path specific datasource for alerting * align endpoint names with cloud * single route per Alerting config * removes unused routing pkg * regens spec * adds datasource param to ruler/prom route paths * Modifications for supporting migration * Apply suggestions from code review * hack for cleaning circular refs in swagger definition * generates files * minor fixes for prom endpoints * decorate prom apis with required: true where applicable * Revert "generates files" This reverts commit ef7e97558477d79bcad416e043b04dbd04a2c8f7. * removes server autogen * Update imported structs from ngalert * Fix listing rules response * Update github.com/prometheus/common dependency * Update get silence response * Update get silences response * adds ruler validation & backend switching * Fix GET /alertmanager/{DatasourceId}/config/api/v1/alerts response * Distinct gettable and postable grafana receivers * Remove permissions routes * Latest JSON specs * Fix testing routes * inline yaml annotation on apirulenode * yaml test & yamlv3 + comments * Fix yaml annotations for embedded type * Rename DatasourceId path parameter * Implement Backend.String() * backend zero value is a real backend * exports DiscoveryBase * Fix GO initialisms * Silences: Use PostableSilence as the base struct for creating silences * Use type alias instead of struct embedding * More fixes to alertmanager silencing routes * post and spec JSONs * Split rule config to postable/gettable * Fix empty POST /silences payload Recreating the generated JSON specs fixes the issue without further modifications * better yaml unmarshaling for nested yaml docs in cortex-am configs * regens spec * re-adds config.receivers * omitempty to align with prometheus API behavior * Prefix routes with /api * Update Alertmanager models * Make adjustments to follow the Alertmanager API * ruler: add for and annotations to grafana alert (#45) * Modify testing API routes * Fix grafana rule for field type * Move PostableUserConfig validation to this library * Fix PostableUserConfig YAML encoding/decoding * Use common fields for grafana and lotex rules * Add namespace id in GettableGrafanaRule * Apply suggestions from code review * fixup * more changes * Apply suggestions from code review * aligns structure pre merge * fix new imports & tests * updates tooling readme * goimports * lint * more linting!! * revive lint Co-authored-by: Sofia Papagiannaki <papagian@gmail.com> Co-authored-by: Domas <domasx2@gmail.com> Co-authored-by: Sofia Papagiannaki <papagian@users.noreply.github.com> Co-authored-by: Ganesh Vernekar <15064823+codesome@users.noreply.github.com> Co-authored-by: gotjosh <josue@grafana.com> Co-authored-by: David Parrott <stomp.box.yo@gmail.com> Co-authored-by: Kyle Brandt <kyle@grafana.com>
2021-04-19 13:26:04 -05:00
apimodels "github.com/grafana/grafana/pkg/services/ngalert/api/tooling/definitions"
ngmodels "github.com/grafana/grafana/pkg/services/ngalert/models"
"github.com/grafana/grafana/pkg/util"
"github.com/prometheus/common/model"
)
type RulerSrv struct {
store store.RuleStore
DatasourceCache datasources.CacheService
manager *state.Manager
log log.Logger
}
func (srv RulerSrv) RouteDeleteNamespaceRulesConfig(c *models.ReqContext) response.Response {
namespaceTitle := c.Params(":Namespace")
namespace, err := srv.store.GetNamespaceByTitle(namespaceTitle, c.SignedInUser.OrgId, c.SignedInUser, true)
if err != nil {
return toNamespaceErrorResponse(err)
}
uids, err := srv.store.DeleteNamespaceAlertRules(c.SignedInUser.OrgId, namespace.Uid)
if err != nil {
return response.Error(http.StatusInternalServerError, "failed to delete namespace alert rules", err)
}
for _, uid := range uids {
srv.manager.RemoveByRuleUID(c.SignedInUser.OrgId, uid)
}
return response.JSON(http.StatusAccepted, util.DynMap{"message": "namespace rules deleted"})
}
func (srv RulerSrv) RouteDeleteRuleGroupConfig(c *models.ReqContext) response.Response {
namespaceTitle := c.Params(":Namespace")
namespace, err := srv.store.GetNamespaceByTitle(namespaceTitle, c.SignedInUser.OrgId, c.SignedInUser, true)
if err != nil {
return toNamespaceErrorResponse(err)
}
ruleGroup := c.Params(":Groupname")
uids, err := srv.store.DeleteRuleGroupAlertRules(c.SignedInUser.OrgId, namespace.Uid, ruleGroup)
if err != nil {
if errors.Is(err, ngmodels.ErrRuleGroupNamespaceNotFound) {
return response.Error(http.StatusNotFound, "failed to delete rule group", err)
}
return response.Error(http.StatusInternalServerError, "failed to delete rule group", err)
}
for _, uid := range uids {
srv.manager.RemoveByRuleUID(c.SignedInUser.OrgId, uid)
}
return response.JSON(http.StatusAccepted, util.DynMap{"message": "rule group deleted"})
}
func (srv RulerSrv) RouteGetNamespaceRulesConfig(c *models.ReqContext) response.Response {
namespaceTitle := c.Params(":Namespace")
namespace, err := srv.store.GetNamespaceByTitle(namespaceTitle, c.SignedInUser.OrgId, c.SignedInUser, false)
if err != nil {
return toNamespaceErrorResponse(err)
}
q := ngmodels.ListNamespaceAlertRulesQuery{
OrgID: c.SignedInUser.OrgId,
NamespaceUID: namespace.Uid,
}
if err := srv.store.GetNamespaceAlertRules(&q); err != nil {
return response.Error(http.StatusInternalServerError, "failed to update rule group", err)
}
result := apimodels.NamespaceConfigResponse{}
ruleGroupConfigs := make(map[string]apimodels.GettableRuleGroupConfig)
for _, r := range q.Result {
ruleGroupConfig, ok := ruleGroupConfigs[r.RuleGroup]
if !ok {
ruleGroupInterval := model.Duration(time.Duration(r.IntervalSeconds) * time.Second)
ruleGroupConfigs[r.RuleGroup] = apimodels.GettableRuleGroupConfig{
Name: r.RuleGroup,
Interval: ruleGroupInterval,
Rules: []apimodels.GettableExtendedRuleNode{
toGettableExtendedRuleNode(*r, namespace.Id),
},
}
} else {
ruleGroupConfig.Rules = append(ruleGroupConfig.Rules, toGettableExtendedRuleNode(*r, namespace.Id))
ruleGroupConfigs[r.RuleGroup] = ruleGroupConfig
}
}
for _, ruleGroupConfig := range ruleGroupConfigs {
result[namespaceTitle] = append(result[namespaceTitle], ruleGroupConfig)
}
return response.JSON(http.StatusAccepted, result)
}
func (srv RulerSrv) RouteGetRulegGroupConfig(c *models.ReqContext) response.Response {
namespaceTitle := c.Params(":Namespace")
namespace, err := srv.store.GetNamespaceByTitle(namespaceTitle, c.SignedInUser.OrgId, c.SignedInUser, false)
if err != nil {
return toNamespaceErrorResponse(err)
}
ruleGroup := c.Params(":Groupname")
q := ngmodels.ListRuleGroupAlertRulesQuery{
OrgID: c.SignedInUser.OrgId,
NamespaceUID: namespace.Uid,
RuleGroup: ruleGroup,
}
if err := srv.store.GetRuleGroupAlertRules(&q); err != nil {
return response.Error(http.StatusInternalServerError, "failed to get group alert rules", err)
}
var ruleGroupInterval model.Duration
ruleNodes := make([]apimodels.GettableExtendedRuleNode, 0, len(q.Result))
for _, r := range q.Result {
ruleGroupInterval = model.Duration(time.Duration(r.IntervalSeconds) * time.Second)
ruleNodes = append(ruleNodes, toGettableExtendedRuleNode(*r, namespace.Id))
}
result := apimodels.RuleGroupConfigResponse{
GettableRuleGroupConfig: apimodels.GettableRuleGroupConfig{
Name: ruleGroup,
Interval: ruleGroupInterval,
Rules: ruleNodes,
},
}
return response.JSON(http.StatusAccepted, result)
}
func (srv RulerSrv) RouteGetRulesConfig(c *models.ReqContext) response.Response {
q := ngmodels.ListAlertRulesQuery{
OrgID: c.SignedInUser.OrgId,
}
if err := srv.store.GetOrgAlertRules(&q); err != nil {
return response.Error(http.StatusInternalServerError, "failed to get alert rules", err)
}
configs := make(map[string]map[string]apimodels.GettableRuleGroupConfig)
for _, r := range q.Result {
folder, err := srv.store.GetNamespaceByUID(r.NamespaceUID, c.SignedInUser.OrgId, c.SignedInUser)
if err != nil {
if errors.Is(err, models.ErrFolderAccessDenied) {
// do not fail if used does not have access to a specific namespace
// just do not include it in the response
continue
}
return toNamespaceErrorResponse(err)
}
namespace := folder.Title
_, ok := configs[namespace]
if !ok {
ruleGroupInterval := model.Duration(time.Duration(r.IntervalSeconds) * time.Second)
configs[namespace] = make(map[string]apimodels.GettableRuleGroupConfig)
configs[namespace][r.RuleGroup] = apimodels.GettableRuleGroupConfig{
Name: r.RuleGroup,
Interval: ruleGroupInterval,
Rules: []apimodels.GettableExtendedRuleNode{
toGettableExtendedRuleNode(*r, folder.Id),
},
}
} else {
ruleGroupConfig, ok := configs[namespace][r.RuleGroup]
if !ok {
ruleGroupInterval := model.Duration(time.Duration(r.IntervalSeconds) * time.Second)
configs[namespace][r.RuleGroup] = apimodels.GettableRuleGroupConfig{
Name: r.RuleGroup,
Interval: ruleGroupInterval,
Rules: []apimodels.GettableExtendedRuleNode{
toGettableExtendedRuleNode(*r, folder.Id),
},
}
} else {
ruleGroupConfig.Rules = append(ruleGroupConfig.Rules, toGettableExtendedRuleNode(*r, folder.Id))
configs[namespace][r.RuleGroup] = ruleGroupConfig
}
}
}
result := apimodels.NamespaceConfigResponse{}
for namespace, m := range configs {
for _, ruleGroupConfig := range m {
result[namespace] = append(result[namespace], ruleGroupConfig)
}
}
return response.JSON(http.StatusAccepted, result)
}
func (srv RulerSrv) RoutePostNameRulesConfig(c *models.ReqContext, ruleGroupConfig apimodels.PostableRuleGroupConfig) response.Response {
namespaceTitle := c.Params(":Namespace")
namespace, err := srv.store.GetNamespaceByTitle(namespaceTitle, c.SignedInUser.OrgId, c.SignedInUser, true)
if err != nil {
return toNamespaceErrorResponse(err)
}
// TODO check permissions
// TODO check quota
// TODO validate UID uniqueness in the payload
//TODO: Should this belong in alerting-api?
if ruleGroupConfig.Name == "" {
return response.Error(http.StatusBadRequest, "rule group name is not valid", nil)
}
for _, r := range ruleGroupConfig.Rules {
cond := ngmodels.Condition{
Condition: r.GrafanaManagedAlert.Condition,
OrgID: c.SignedInUser.OrgId,
Data: r.GrafanaManagedAlert.Data,
}
if err := validateCondition(cond, c.SignedInUser, c.SkipCache, srv.DatasourceCache); err != nil {
return response.Error(http.StatusBadRequest, fmt.Sprintf("failed to validate alert rule %s", r.GrafanaManagedAlert.Title), err)
}
}
if err := srv.store.UpdateRuleGroup(store.UpdateRuleGroupCmd{
OrgID: c.SignedInUser.OrgId,
NamespaceUID: namespace.Uid,
RuleGroupConfig: ruleGroupConfig,
}); err != nil {
if errors.Is(err, ngmodels.ErrAlertRuleNotFound) {
return response.Error(http.StatusNotFound, "failed to update rule group", err)
} else if errors.Is(err, ngmodels.ErrAlertRuleFailedValidation) {
return response.Error(http.StatusBadRequest, "failed to update rule group", err)
}
return response.Error(http.StatusInternalServerError, "failed to update rule group", err)
}
return response.JSON(http.StatusAccepted, util.DynMap{"message": "rule group updated successfully"})
}
func toGettableExtendedRuleNode(r ngmodels.AlertRule, namespaceID int64) apimodels.GettableExtendedRuleNode {
gettableExtendedRuleNode := apimodels.GettableExtendedRuleNode{
GrafanaManagedAlert: &apimodels.GettableGrafanaRule{
ID: r.ID,
OrgID: r.OrgID,
Title: r.Title,
Condition: r.Condition,
Data: r.Data,
Updated: r.Updated,
IntervalSeconds: r.IntervalSeconds,
Version: r.Version,
UID: r.UID,
NamespaceUID: r.NamespaceUID,
NamespaceID: namespaceID,
RuleGroup: r.RuleGroup,
NoDataState: apimodels.NoDataState(r.NoDataState),
ExecErrState: apimodels.ExecutionErrorState(r.ExecErrState),
},
}
gettableExtendedRuleNode.ApiRuleNode = &apimodels.ApiRuleNode{
For: model.Duration(r.For),
Annotations: r.Annotations,
Labels: r.Labels,
}
return gettableExtendedRuleNode
}
func toNamespaceErrorResponse(err error) response.Response {
if errors.Is(err, ngmodels.ErrCannotEditNamespace) {
return response.Error(http.StatusForbidden, err.Error(), err)
}
if errors.Is(err, models.ErrDashboardIdentifierNotSet) {
return response.Error(http.StatusBadRequest, err.Error(), err)
}
return coreapi.ToFolderErrorResponse(err)
}