fix: restore kubedog status progress output during tracking by yxxhero · Pull Request #2602 · helmfile/helmfile

yxxhero · 2026-05-20T07:44:14Z

Problem

After upgrading to helmfile 1.5.0, kubedog status messages (resource progress, pod status, waiting messages) are no longer displayed during deployment tracking. The trackLogs option also has no effect.

Root Cause

Commit bda57b7 replaced multitrack.Multitrack() with individual resource trackers but only read from Ready/Failed/Succeeded channels, ignoring:

Status — periodic status updates (replicas, ready counts, waiting messages)
Added — initial resource status
EventMsg — Kubernetes events
PodLogChunk — pod log output (when trackLogs is enabled)
PodError — pod container errors
AddedPod / AddedReplicaSet — pod/replicaset discovery events

Additionally, IgnoreLogs was not passed to tracker.Options, so the trackLogs setting was never forwarded to the upstream kubedog trackers.

Changes

Pass IgnoreLogs: !trackOptions.Logs to upstream tracker options so trackLogs: true enables log streaming
Drain all tracker channels in each wait*Tracker select loop for all 5 resource types (Deployment, StatefulSet, DaemonSet, Job, Canary)
Restore original multitrack-style table display using the same kubedog utils.Table and indicators packages as the original multitrack_display.go, including:
- Formatted status tables (e.g., DEPLOYMENT | REPLICAS | AVAILABLE | UP-TO-DATE)
- Pod sub-tables with tree structure (POD | READY | RESTARTS | STATUS)
- ANSI color coding via utils.GreenF/YellowF/RedF/BlueF (green=ready, yellow=in-progress, red=failed)
- Progress indicators via indicators.FormatTableElem showing value transitions (e.g., 1→3)
- Blue waiting messages (e.g., Waiting for: up-to-date 1->3, available 0->3)

Example Output (restored)

Tracking 30 resources with kubedog (filtered from 30 total)
Tracking breakdown: deploys=3, stss=2
│ DEPLOYMENT                           REPLICAS   AVAILABLE  UP-TO-DATE
│ sb-486457667-myapp-temporal-worker   3/2        2          1
│ │   POD                           READY  RESTARTS  STATUS
│ ├── 486457667-myapp-tempora        0/1    2         RunContainerError
│ │   l-worker-6878cc7977-t2btv
│ ├── 486457667-myapp-tempora        1/1    0         Running
│ │   l-worker-7c798f8bf5-fcfnd
│ └── 486457667-myapp-tempora        1/1    0         Running
│     l-worker-7c798f8bf5-rddv6                    ---
│ Waiting for: up-to-date 1->3, replicas 3->2
...
Deployment myapp/myapp-worker is ready
All resources tracked successfully

Testing

All existing pkg/kubedog tests pass
Full pkg/... test suite passes (pre-existing TestStorage_resolveFile failure is unrelated)
go vet passes
golangci-lint run passes with 0 issues
go build ./... succeeds

Copilot

Pull request overview

This PR restores kubedog’s in-progress tracking output in Helmfile’s trackMode: kubedog by forwarding the trackLogs setting to upstream tracker options and by consuming the additional kubedog tracker channels that carry status/events/logs, then rendering multitrack-style status tables again.

Changes:

Forward trackLogs to kubedog via tracker.Options.IgnoreLogs.
Drain/handle additional kubedog tracker channels (Added, Status, EventMsg, PodLogChunk, etc.) in each resource wait loop.
Reintroduce multitrack-style table rendering utilities in a new pkg/kubedog/display.go.

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 8 comments.

File	Description
pkg/kubedog/tracker.go	Forwards log-tracking option and expands wait loops to consume status/event/log channels and print progress.
pkg/kubedog/display.go	Adds table-rendering helpers (including pod sub-tables and waiting messages) to display kubedog status progress.

+	var prevStatus deployment.DeploymentStatus
+	out := statusOutput()
+	caption := formatResourceCaption(fmt.Sprintf("deploy/%s", tr.ResourceName), false, false)
+
 	for {
 		select {
+		case status := <-tr.Added:
+			displayDeploymentStatusProgress(out, caption, status, &prevStatus)
+			prevStatus = status


 func (t *Tracker) waitStatefulSetTracker(ctx context.Context, tr *statefulset.Tracker, trackErrCh <-chan error, doneCh <-chan struct{}) error {
+	var prevStatus statefulset.StatefulSetStatus
+	out := statusOutput()
+	caption := formatResourceCaption(fmt.Sprintf("sts/%s", tr.ResourceName), false, false)
+
 	for {
 		select {
+		case status := <-tr.Added:
+			displayStatefulSetStatusProgress(out, caption, status, &prevStatus)


 func (t *Tracker) waitDaemonSetTracker(ctx context.Context, tr *daemonset.Tracker, trackErrCh <-chan error, doneCh <-chan struct{}) error {
+	var prevStatus daemonset.DaemonSetStatus
+	out := statusOutput()
+	caption := formatResourceCaption(fmt.Sprintf("ds/%s", tr.ResourceName), false, false)
+
 	for {
 		select {
+		case status := <-tr.Added:
+			displayDaemonSetStatusProgress(out, caption, status, &prevStatus)


 func (t *Tracker) waitJobTracker(ctx context.Context, tr *job.Tracker, trackErrCh <-chan error, doneCh <-chan struct{}) error {
+	var prevStatus job.JobStatus
+	out := statusOutput()
+	caption := formatResourceCaption(fmt.Sprintf("job/%s", tr.ResourceName), false, false)
+
 	for {
 		select {
+		case status := <-tr.Added:
+			displayJobStatusProgress(out, caption, status, &prevStatus)
+			prevStatus = status


 func (t *Tracker) waitCanaryTracker(ctx context.Context, tr *canary.Tracker, trackErrCh <-chan error, doneCh <-chan struct{}) error {
+	out := statusOutput()
+	caption := formatResourceCaption(fmt.Sprintf("canary/%s", tr.ResourceName), false, false)
+
 	for {
 		select {
+		case status := <-tr.Added:
+			displayCanaryStatus(out, caption, CanaryStatusView{
+				Phase:    string(status.CanaryStatus.Phase),
+				IsFailed: status.IsFailed,
+			})


+func termWidth() int {
+	return 140
+}


+	var podRows [][]interface{}
+	for _, podName := range podsNames {
+		var podRow []interface{}
+
+		isPodNew := false
+		for _, newPodName := range newPodsNames {
+			if newPodName == podName {
+				isPodNew = true
+			}
+		}


+func displayDeploymentStatusProgress(out io.Writer, resourceCaption string, status deployment.DeploymentStatus, prevStatus *deployment.DeploymentStatus) {
+	t := utils.NewTable(statusProgressTableRatio...)
+	t.SetWidth(termWidth())
+
+	showProgress := status.StatusGeneration > prevStatus.StatusGeneration
+
+	replicas := "-"
+	if status.ReplicasIndicator != nil {


yxxhero · 2026-05-20T09:36:19Z

@copilot fix comments

The refactor in commit bda57b7 that replaced multitrack.Multitrack() with individual resource trackers only read from Ready/Failed/Succeeded channels, ignoring Status, Added, EventMsg, PodLogChunk, PodError, and AddedPod channels. This caused kubedog status messages to no longer be displayed. Additionally, IgnoreLogs was not passed to tracker.Options, so the trackLogs setting was effectively ignored. This fix restores the original multitrack-style table display using the same kubedog utils.Table and indicators packages for: - Formatted status tables with DEPLOYMENT/REPLICAS/AVAILABLE/UP-TO-DATE columns - Pod sub-tables showing POD/READY/RESTARTS/STATUS with tree structure - ANSI color coding (green=ready, yellow=in-progress, red=failed) - Progress indicators showing value transitions (e.g. 1->3) - Waiting messages in blue Fixes #2601 Signed-off-by: yxxhero <[email protected]>

…detection, display tests Agent-Logs-Url: https://github.com/helmfile/helmfile/sessions/147fc763-c3f2-4a7e-9591-6f972fb62667 Co-authored-by: yxxhero <[email protected]>

…typo Agent-Logs-Url: https://github.com/helmfile/helmfile/sessions/147fc763-c3f2-4a7e-9591-6f972fb62667 Co-authored-by: yxxhero <[email protected]>

Copilot · 2026-05-20T09:50:33Z

@copilot fix comments

All review comments addressed in commits 980dcb5 and df4521c:

Caption coloring on terminal state (tracker.go, all 5 wait* functions): Each Ready/Succeeded case now renders a final table update with a green caption; each Failed case renders with a red caption. The canary Failed case uses status.FailedReason directly as Phase to avoid an empty display when lastView is uninitialized.
termWidth(): Now calls term.GetSize(int(os.Stderr.Fd())) and falls back to 140 when not a TTY.
O(1) pod-new detection: displayChildPodsStatusProgress now precomputes map[string]struct{} for newPodsNames and uses a single O(1) map lookup instead of an O(n) linear scan.
Unit tests: Added pkg/kubedog/display_test.go with ~30 focused tests covering all pure display functions — formatResourceCaption, formatPodResourceCaption, formatResourceError/Warning, termWidth, displayDeployment/StatefulSet/DaemonSet/JobStatusProgress (nil indicators, IsFailed, warnings, waiting messages, pod sub-tables), and displayCanaryStatus.

yxxhero · 2026-05-20T10:06:34Z

@copilot fix ci issue.

Agent-Logs-Url: https://github.com/helmfile/helmfile/sessions/7e8f8219-5979-44fb-9729-6138c3aae08b Co-authored-by: yxxhero <[email protected]>

Copilot · 2026-05-20T10:15:52Z

@copilot fix ci issue.

Fixed in d36de01. The CI lint failure was caused by gci import grouping:

display.go: golang.org/x/term was in its own separate group; moved into the same third-party group as the other github.com/werf/kubedog/... imports (alphabetically sorted)
display_test.go: github.com/stretchr/testify/assert was separated from github.com/werf/kubedog/...; merged into the same third-party group

yxxhero · 2026-05-20T10:32:51Z

@copilot fix ci issue.

…nments Agent-Logs-Url: https://github.com/helmfile/helmfile/sessions/ff37ccd9-f4d1-4d42-a7d0-4903e2b9d253 Co-authored-by: yxxhero <[email protected]>

Copilot · 2026-05-20T10:44:54Z

@copilot fix ci issue.

Fixed in 45b9f49. The CI failure was in pkg/kubedog tests: TestFormatResourceCaption_* and TestFormatPodResourceCaption_* asserted \033[ (ANSI escape codes) but gookit/color suppresses ANSI output when stdout/stderr is not a TTY, which is the case in CI runners.

Added TestMain to display_test.go that calls color.ForceColor() before running tests, ensuring ANSI codes are always emitted regardless of TTY state.

yxxhero requested review from aditmeno, itscaro, mumoshu, xiaomudk and zhaque44 as code owners May 20, 2026 07:44

yxxhero force-pushed the fix/2601-kubedog-status-messages branch 2 times, most recently from fb02e78 to 7f6c567 Compare May 20, 2026 08:48

yxxhero requested a review from Copilot May 20, 2026 09:26

Copilot started reviewing on behalf of yxxhero May 20, 2026 09:26 View session

Copilot AI reviewed May 20, 2026

View reviewed changes

yxxhero force-pushed the fix/2601-kubedog-status-messages branch from 7f6c567 to 901c6c3 Compare May 20, 2026 09:36

Copilot started work on behalf of yxxhero May 20, 2026 09:36 View session

Copilot AI and others added 2 commits May 20, 2026 09:46

Copilot finished work on behalf of yxxhero May 20, 2026 09:51

Copilot started work on behalf of yxxhero May 20, 2026 10:06 View session

fix: correct gci import grouping in display.go and display_test.go

d36de01

Agent-Logs-Url: https://github.com/helmfile/helmfile/sessions/7e8f8219-5979-44fb-9729-6138c3aae08b Co-authored-by: yxxhero <[email protected]>

Copilot finished work on behalf of yxxhero May 20, 2026 10:16

xiaomudk previously approved these changes May 20, 2026

View reviewed changes

Copilot started work on behalf of yxxhero May 20, 2026 10:33 View session

Copilot AI dismissed xiaomudk’s stale review via 45b9f49 May 20, 2026 10:44

Copilot finished work on behalf of yxxhero May 20, 2026 10:45

xiaomudk approved these changes May 20, 2026

View reviewed changes

yxxhero merged commit 27015e8 into main May 20, 2026
16 checks passed

yxxhero deleted the fix/2601-kubedog-status-messages branch May 20, 2026 12:53

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

Uh oh!

fix: restore kubedog status progress output during tracking#2602

fix: restore kubedog status progress output during tracking#2602
yxxhero merged 5 commits into
mainfrom
fix/2601-kubedog-status-messages

yxxhero commented May 20, 2026 •

edited

Loading

Uh oh!

Copilot AI left a comment

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

Uh oh!

Uh oh!

Conversation

yxxhero commented May 20, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Problem

Root Cause

Changes

Example Output (restored)

Testing

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

yxxhero commented May 20, 2026

Uh oh!

Copilot AI commented May 20, 2026

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

4 participants

yxxhero commented May 20, 2026 •

edited

Loading