Show a target paused by its circuit breaker (closes #385)
check / check (push) Successful in 3m21s
check / check (push) Successful in 3m21s
While an http or slack target's circuit breaker turns its deliveries away, the target's row on the webhook page says its deliveries are paused and until when, in UTC and from now. Each of its retrying deliveries shows as waiting, with the reason and the same time, in the event log and on the event's page, instead of a bare "retrying". The engine gains one read, CooldownRemaining(targetID), over the breakers it already keeps; the handlers reach it through a one-method interface wired like Archives. Model: opus-5-5
This commit is contained in:
@@ -57,19 +57,20 @@ var errVerificationBusy = errors.New(
|
||||
type HandlersParams struct {
|
||||
fx.In
|
||||
|
||||
Logger *logger.Logger
|
||||
Globals *globals.Globals
|
||||
Config *config.Config
|
||||
Database *database.Database
|
||||
WebhookDBMgr *database.WebhookDBManager
|
||||
Healthcheck *healthcheck.Healthcheck
|
||||
Session *session.Session
|
||||
Middleware *middleware.Middleware
|
||||
Notifier delivery.Notifier
|
||||
Archives delivery.Archives
|
||||
SSRFGuard *delivery.Guard
|
||||
Metrics *metrics.Set
|
||||
Registry *prometheus.Registry
|
||||
Logger *logger.Logger
|
||||
Globals *globals.Globals
|
||||
Config *config.Config
|
||||
Database *database.Database
|
||||
WebhookDBMgr *database.WebhookDBManager
|
||||
Healthcheck *healthcheck.Healthcheck
|
||||
Session *session.Session
|
||||
Middleware *middleware.Middleware
|
||||
Notifier delivery.Notifier
|
||||
Archives delivery.Archives
|
||||
CircuitBreakers delivery.CircuitBreakers
|
||||
SSRFGuard *delivery.Guard
|
||||
Metrics *metrics.Set
|
||||
Registry *prometheus.Registry
|
||||
}
|
||||
|
||||
// Handlers provides HTTP handler methods for all application
|
||||
@@ -84,6 +85,7 @@ type Handlers struct {
|
||||
mw *middleware.Middleware
|
||||
notifier delivery.Notifier
|
||||
archives delivery.Archives
|
||||
breakers delivery.CircuitBreakers
|
||||
mtr *metrics.Set
|
||||
templates map[string]*template.Template
|
||||
|
||||
@@ -145,6 +147,7 @@ func New(
|
||||
s.mw = params.Middleware
|
||||
s.notifier = params.Notifier
|
||||
s.archives = params.Archives
|
||||
s.breakers = params.CircuitBreakers
|
||||
s.mtr = params.Metrics
|
||||
s.ssrf = params.SSRFGuard
|
||||
|
||||
|
||||
@@ -9,6 +9,7 @@ import (
|
||||
"net/http/httptest"
|
||||
"sync"
|
||||
"testing"
|
||||
"time"
|
||||
|
||||
"github.com/stretchr/testify/assert"
|
||||
"github.com/stretchr/testify/require"
|
||||
@@ -181,6 +182,47 @@ func (r *recordingArchives) Renames() []archiveRename {
|
||||
return out
|
||||
}
|
||||
|
||||
// testCircuitBreakers is a delivery.CircuitBreakers over real circuit
|
||||
// breakers, one per target a test asks for, so a test can trip a
|
||||
// target's breaker with RecordFailure and close it with RecordSuccess.
|
||||
type testCircuitBreakers struct {
|
||||
mu sync.Mutex
|
||||
breakers map[string]*delivery.CircuitBreaker
|
||||
}
|
||||
|
||||
// Breaker returns the target's circuit breaker, making it on first
|
||||
// use.
|
||||
func (b *testCircuitBreakers) Breaker(
|
||||
targetID string,
|
||||
) *delivery.CircuitBreaker {
|
||||
b.mu.Lock()
|
||||
defer b.mu.Unlock()
|
||||
|
||||
if b.breakers == nil {
|
||||
b.breakers = map[string]*delivery.CircuitBreaker{}
|
||||
}
|
||||
|
||||
if b.breakers[targetID] == nil {
|
||||
b.breakers[targetID] = delivery.NewCircuitBreaker()
|
||||
}
|
||||
|
||||
return b.breakers[targetID]
|
||||
}
|
||||
|
||||
func (b *testCircuitBreakers) CooldownRemaining(
|
||||
targetID string,
|
||||
) time.Duration {
|
||||
b.mu.Lock()
|
||||
defer b.mu.Unlock()
|
||||
|
||||
cb := b.breakers[targetID]
|
||||
if cb == nil {
|
||||
return 0
|
||||
}
|
||||
|
||||
return cb.CooldownRemaining()
|
||||
}
|
||||
|
||||
// newTestApp returns an app whose RequireStart fails the test when
|
||||
// starting takes longer than fx's default start timeout of 15s. That
|
||||
// limit catches a start that hangs, not a busy host: measured with make
|
||||
@@ -231,6 +273,12 @@ func newTestAppWithConfig(
|
||||
func(r *recordingArchives) delivery.Archives {
|
||||
return r
|
||||
},
|
||||
func() *testCircuitBreakers {
|
||||
return &testCircuitBreakers{}
|
||||
},
|
||||
func(b *testCircuitBreakers) delivery.CircuitBreakers {
|
||||
return b
|
||||
},
|
||||
metrics.NewRegistry,
|
||||
metrics.New,
|
||||
middleware.New,
|
||||
|
||||
@@ -109,6 +109,11 @@ type DeliveryView struct {
|
||||
// the middle of Results. The page must show it, or the
|
||||
// bound would hide history rather than fold it.
|
||||
AttemptsOmitted int
|
||||
|
||||
// Paused is set while the delivery is retrying and its
|
||||
// target's circuit breaker is turning deliveries away, and
|
||||
// nil otherwise.
|
||||
Paused *PausedView
|
||||
}
|
||||
|
||||
// eventLogTarget is what the event log needs to know about
|
||||
@@ -1181,7 +1186,7 @@ func (h *Handlers) eventLogViews(
|
||||
}
|
||||
|
||||
for i := range rows {
|
||||
result[i].Deliveries = newDeliveryViews(
|
||||
result[i].Deliveries = h.newDeliveryViews(
|
||||
eventDeliveries[i], targetMap, attempts,
|
||||
)
|
||||
result[i].ResubmitCount = resubmits[rows[i].ID]
|
||||
@@ -1305,8 +1310,9 @@ func (h *Handlers) loadDeliveryResults(
|
||||
|
||||
// newDeliveryViews projects deliveries for rendering,
|
||||
// resolving each one's target to its display-safe view and
|
||||
// each one's attempts through that target's redactor.
|
||||
func newDeliveryViews(
|
||||
// each one's attempts through that target's redactor. A
|
||||
// retrying delivery also reads its target's circuit breaker.
|
||||
func (h *Handlers) newDeliveryViews(
|
||||
deliveries []database.Delivery,
|
||||
targetMap map[string]eventLogTarget,
|
||||
attempts map[string][]deliveryResultRow,
|
||||
@@ -1329,6 +1335,10 @@ func newDeliveryViews(
|
||||
AttemptCount: len(rows),
|
||||
AttemptsOmitted: omitted,
|
||||
}
|
||||
|
||||
if deliveries[i].Status == database.DeliveryStatusRetrying {
|
||||
views[i].Paused = h.pausedView(deliveries[i].TargetID)
|
||||
}
|
||||
}
|
||||
|
||||
return views
|
||||
|
||||
@@ -24,6 +24,34 @@ type TargetRowView struct {
|
||||
// Archive is a database target's archive file, and nil for a target
|
||||
// of any other type.
|
||||
Archive *ArchiveFileView
|
||||
|
||||
// Paused is set while the target's circuit breaker is turning its
|
||||
// deliveries away, and nil otherwise.
|
||||
Paused *PausedView
|
||||
}
|
||||
|
||||
// PausedView is when a target's circuit breaker will let deliveries to
|
||||
// it through again: Until is the time in UTC, and Relative how long
|
||||
// that is from now.
|
||||
type PausedView struct {
|
||||
Until string
|
||||
Relative string
|
||||
}
|
||||
|
||||
// pausedView reads the target's circuit breaker, and returns nil when
|
||||
// the breaker is not turning the target's deliveries away.
|
||||
func (h *Handlers) pausedView(targetID string) *PausedView {
|
||||
remaining := h.breakers.CooldownRemaining(targetID)
|
||||
if remaining <= 0 {
|
||||
return nil
|
||||
}
|
||||
|
||||
until := time.Now().Add(remaining)
|
||||
|
||||
return &PausedView{
|
||||
Until: until.UTC().Format(time.TimeOnly) + " UTC",
|
||||
Relative: humanize.Time(until),
|
||||
}
|
||||
}
|
||||
|
||||
// TargetDeliveries is how many of a target's deliveries became
|
||||
@@ -84,6 +112,8 @@ func (h *Handlers) targetRows(
|
||||
if targets[i].Type == database.TargetTypeDatabase {
|
||||
rows[i].Archive = h.archiveFileView(webhook, &targets[i])
|
||||
}
|
||||
|
||||
rows[i].Paused = h.pausedView(targets[i].ID)
|
||||
}
|
||||
|
||||
return rows
|
||||
|
||||
@@ -0,0 +1,129 @@
|
||||
package handlers_test
|
||||
|
||||
import (
|
||||
"net/http"
|
||||
"strings"
|
||||
"testing"
|
||||
|
||||
"github.com/stretchr/testify/assert"
|
||||
"github.com/stretchr/testify/require"
|
||||
"gorm.io/gorm/clause"
|
||||
"sneak.berlin/go/webhooker/internal/database"
|
||||
"sneak.berlin/go/webhooker/internal/delivery"
|
||||
"sneak.berlin/go/webhooker/internal/handlers"
|
||||
"sneak.berlin/go/webhooker/internal/session"
|
||||
)
|
||||
|
||||
// resumesAt is how the pages write when a paused target's deliveries
|
||||
// resume: the time in UTC, then how long that is from now.
|
||||
const resumesAt = `\d\d:\d\d:\d\d UTC \(\d+ seconds from now\)`
|
||||
|
||||
// TestPausedTarget_ShownUntilBreakerCloses trips the circuit breaker of
|
||||
// an http target, then checks that its row on the webhook page says its
|
||||
// deliveries are paused and until when, and that its retrying delivery
|
||||
// says it is waiting and why in the event log and on the event's page,
|
||||
// while its delivered delivery and the log target are shown as before.
|
||||
// Once the breaker closes, none of the pages says so any more.
|
||||
func TestPausedTarget_ShownUntilBreakerCloses(t *testing.T) {
|
||||
t.Parallel()
|
||||
|
||||
var (
|
||||
h *handlers.Handlers
|
||||
sess *session.Session
|
||||
db *database.Database
|
||||
dbMgr *database.WebhookDBManager
|
||||
breakers *testCircuitBreakers
|
||||
)
|
||||
|
||||
app := newTestApp(t, &h, &sess, &db, &dbMgr, &breakers)
|
||||
app.RequireStart()
|
||||
|
||||
t.Cleanup(app.RequireStop)
|
||||
|
||||
wh := seedWebhook(t, db)
|
||||
target := seedTarget(t, db, wh.ID, database.TargetTypeHTTP)
|
||||
seedTarget(t, db, wh.ID, database.TargetTypeLog)
|
||||
|
||||
retrying := seedStoredEvent(t, dbMgr, wh.ID, `{"n":1}`)
|
||||
addDelivery(t, dbMgr, wh.ID, retrying.ID, target.ID,
|
||||
database.DeliveryStatusRetrying)
|
||||
|
||||
delivered := seedStoredEvent(t, dbMgr, wh.ID, `{"n":2}`)
|
||||
addDelivery(t, dbMgr, wh.ID, delivered.ID, target.ID,
|
||||
database.DeliveryStatusDelivered)
|
||||
|
||||
cb := breakers.Breaker(target.ID)
|
||||
for cb.State() != delivery.CircuitOpen {
|
||||
cb.RecordFailure()
|
||||
}
|
||||
|
||||
const waiting = "waiting: target paused after repeated failures, " +
|
||||
"resumes " + resumesAt
|
||||
|
||||
list := targetList(t, renderSourceDetailPage(t, h, sess, wh.ID))
|
||||
assert.Regexp(t, "t-http http Active Edit Deactivate Delete "+
|
||||
"Deliveries Paused: after repeated failures, until "+resumesAt,
|
||||
list)
|
||||
assert.Equal(t, 1, strings.Count(list, "Paused"))
|
||||
|
||||
log := renderSourceLogsPage(t, h, sess, wh.ID)
|
||||
assert.Contains(t, log, "t-http: waiting")
|
||||
assert.Contains(t, log, "t-http: delivered")
|
||||
assert.Regexp(t, waiting, log)
|
||||
assert.NotContains(t, log, "retrying")
|
||||
|
||||
page := eventPage(t, h, sess, wh.ID, retrying.ID)
|
||||
assert.Regexp(t, waiting, page)
|
||||
assert.NotContains(t, page, "retrying")
|
||||
|
||||
cb.RecordSuccess()
|
||||
|
||||
list = targetList(t, renderSourceDetailPage(t, h, sess, wh.ID))
|
||||
assert.NotContains(t, list, "Paused")
|
||||
|
||||
log = renderSourceLogsPage(t, h, sess, wh.ID)
|
||||
assert.Contains(t, log, "t-http: retrying")
|
||||
assert.NotContains(t, log, "waiting")
|
||||
|
||||
page = eventPage(t, h, sess, wh.ID, retrying.ID)
|
||||
assert.Contains(t, page, ">retrying</span>")
|
||||
assert.NotContains(t, page, "waiting")
|
||||
}
|
||||
|
||||
// addDelivery records a delivery of the event to the target, with the
|
||||
// given status, in the webhook's own database.
|
||||
func addDelivery(
|
||||
t *testing.T,
|
||||
dbMgr *database.WebhookDBManager,
|
||||
webhookID, eventID, targetID string,
|
||||
status database.DeliveryStatus,
|
||||
) {
|
||||
t.Helper()
|
||||
|
||||
webhookDB, err := dbMgr.GetDB(webhookID)
|
||||
require.NoError(t, err)
|
||||
|
||||
require.NoError(t, webhookDB.Omit(clause.Associations).Create(
|
||||
&database.Delivery{
|
||||
EventID: eventID,
|
||||
TargetID: targetID,
|
||||
Status: status,
|
||||
},
|
||||
).Error)
|
||||
}
|
||||
|
||||
// eventPage runs the real event page handler and returns the
|
||||
// rendered HTML.
|
||||
func eventPage(
|
||||
t *testing.T,
|
||||
h *handlers.Handlers,
|
||||
sess *session.Session,
|
||||
webhookID, eventID string,
|
||||
) string {
|
||||
t.Helper()
|
||||
|
||||
w := serveEventPage(t, h, sess, webhookID, eventID)
|
||||
require.Equal(t, http.StatusOK, w.Code)
|
||||
|
||||
return w.Body.String()
|
||||
}
|
||||
Reference in New Issue
Block a user