Compare commits

...
Author SHA1 Message Date
mrhid6 a0fbf5b9ba fix: Word and colour Windows workloads correctly across the UI
Server Deploy / deploy (push) Canceled after 0s
Chart Release / chart (push) Canceled after 0s
Agent Release / build (push) Successful in 14m1s
Agent Release / msi (push) Successful in 1m22s
2026-08-13 12:57:37 +00:00
mrhid6 ddf0814803 docs: Fix Windows package-inventory and poll-loop claims in CLAUDE.md 2026-08-13 12:28:51 +00:00
10 changed files with 71 additions and 25 deletions
+11 -7
View File
@@ -383,9 +383,11 @@ one wire shape, worded per platform in the UI, which is the only layer that
knows the host's OS. The platform split lives entirely in the agent, as build
tags (`systemd_linux.go` / `services_windows.go` and the matching `control_`
and `logs_` pairs); the control plane is OS-blind and needed no changes.
Windows collection runs PowerShell through `agent/internal/winexec`, and every
script emits JSON that a build-tag-free parser reads, so the parsers are tested
on Linux — the agent module has no Windows CI.
Windows collection runs PowerShell through `agent/internal/winexec`. Every
script that reports data emits JSON that a build-tag-free parser reads, so
those parsers are tested on Linux — the agent module has no Windows CI. The
control verbs and `serviceDisplayName` emit no JSON and have no parser; they
are exercised only by running the agent on Windows.
**Not gated by licence**: this reads as core fleet management, so v1 ships
everywhere with no `HasFeature` check. If that changes the check belongs at
@@ -895,7 +897,7 @@ tls: true
```
1. SyncKeys(server_id, agent_token, agent_version)
2. Non-Linux hosts stop here — Windows agents register and heartbeat only
2. Non-Linux hosts stop here — the key-management steps below are Linux-only; a Windows agent's other work (workflow steps, inventory, OS updates, workloads) runs from the goroutines started above, not from this loop
3. Diff desired keys against /root/.ssh/authorized_keys; unchanged → no write
4. Changed → write .tmp, os.Rename() over the real file, chmod 0600
```
@@ -1171,9 +1173,11 @@ git push origin main # server + web deploy
- **Windows agents cover the fleet-management path** — register, heartbeat, run
steps, report inventory, OS updates through the Windows Update COM API, and
workloads (services plus containers, with control and logs). They still do no
`authorized_keys` management, and no package inventory or CVE matching: the
vulnerability feeds this project uses carry no Windows data, so a Windows host
correctly reports `unsupported` rather than a clean bill of health.
`authorized_keys` management, and no package inventory or CVE matching: a
Windows agent never calls `ReportPackages`, so no `server_packages` document
exists for it and it reports no package inventory at all — a different,
earlier state than the `unsupported` a Linux distribution reaches when its
family has no security feed.
- **Both `server` and `web` scale horizontally** — see "Running more than one server replica" below. `web` holds nothing; `server` holds per-agent state that is routed between replicas over Redis rather than duplicated.
- **Deletion lives in the control plane** — admin sends the warnings because it knows the billing address; the control plane performs the delete because it is the only service that knows which collections carry `instance_id`. Mirroring that list into admin would drift, and a drift there deletes the wrong rows.
+7 -3
View File
@@ -19,12 +19,16 @@ func Run(ctx context.Context, script string) (string, error) {
out, err := cmd.Output()
if err != nil {
if ee, ok := err.(*exec.ExitError); ok && len(ee.Stderr) > 0 {
return "", fmt.Errorf("powershell: %s", strings.TrimSpace(string(ee.Stderr)))
}
// Checked before the ExitError/stderr branch: CommandContext kills the
// process on timeout, and that kill can itself produce an ExitError
// carrying stderr text, so a genuine timeout would otherwise surface
// as that stderr instead of the "timed out" message callers match on.
if ctx.Err() == context.DeadlineExceeded {
return "", fmt.Errorf("powershell: timed out")
}
if ee, ok := err.(*exec.ExitError); ok && len(ee.Stderr) > 0 {
return "", fmt.Errorf("powershell: %s", strings.TrimSpace(string(ee.Stderr)))
}
return "", fmt.Errorf("powershell: %w", err)
}
return string(out), nil
+10 -2
View File
@@ -12,8 +12,16 @@ const servicesTimeout = 60 * time.Second
const servicesScript = `
$ErrorActionPreference = 'Stop'
$svcs = Get-CimInstance Win32_Service |
Select-Object Name,DisplayName,State,StartMode,PathName,ExitCode
$svcs = Get-CimInstance Win32_Service | ForEach-Object {
[pscustomobject]@{
Name = $_.Name
DisplayName = $_.DisplayName
State = $_.State
StartMode = $_.StartMode
PathName = $_.PathName
ExitCode = $_.ExitCode
}
}
ConvertTo-Json -InputObject @($svcs) -Depth 3 -Compress
`
+11 -3
View File
@@ -113,10 +113,15 @@ func parseServices(jsonText, systemRoot string) ([]Workload, error) {
continue
}
state := "stopped"
// The wire shape is shared with the systemd collector — both report
// under kind "unit" — so the state word has to be too, or the UI
// (which colours and filters on it, and does so before it knows
// which platform sent the row) needs two vocabularies for one kind.
// running/stopped/failed become active/inactive/failed to match.
state := "inactive"
switch {
case running:
state = "running"
state = "active"
case failed:
state = "failed"
}
@@ -189,7 +194,10 @@ func parseEvents(jsonText, serviceName, displayName string, tail int) (string, e
continue
}
}
msg := strings.TrimSpace(strings.ReplaceAll(e.M, "\r\n", " "))
// Collapse every newline form, not just "\r\n": a message containing a
// bare "\n" would otherwise still break the one-line-per-event shape
// this renders for the log dialog, and undercount the tail trim above.
msg := strings.TrimSpace(strings.NewReplacer("\r\n", " ", "\r", " ", "\n", " ").Replace(e.M))
lines = append(lines, e.T+" "+e.L+" "+msg)
}
+25 -6
View File
@@ -59,12 +59,12 @@ func TestParseServicesFilters(t *testing.T) {
t.Fatalf("got %d workloads, want 3: %+v", len(got), got)
}
if w := byID["Contoso"]; w.Kind != "unit" || w.Name != "Contoso Broker" || w.State != "running" {
if w := byID["Contoso"]; w.Kind != "unit" || w.Name != "Contoso Broker" || w.State != "active" {
t.Errorf("Contoso = %+v", w)
}
// Enabled but not running is exactly the row worth seeing.
if byID["Fabrikam"].State != "stopped" {
t.Errorf("Fabrikam state = %q, want stopped", byID["Fabrikam"].State)
if byID["Fabrikam"].State != "inactive" {
t.Errorf("Fabrikam state = %q, want inactive", byID["Fabrikam"].State)
}
// A non-zero exit code on a stopped service is a crash, not a clean stop.
if byID["Crashed"].State != "failed" {
@@ -80,15 +80,15 @@ func TestParseServicesExitCode1077(t *testing.T) {
if err != nil {
t.Fatalf("parseServices: %v", err)
}
if len(got) != 1 || got[0].State != "stopped" {
t.Fatalf("got %+v, want one stopped workload", got)
if len(got) != 1 || got[0].State != "inactive" {
t.Fatalf("got %+v, want one inactive workload", got)
}
}
func TestParseServicesSingleObjectAndEmpty(t *testing.T) {
one := `{"Name":"Solo","DisplayName":"Solo","State":"Running","StartMode":"Auto","PathName":"C:\\Solo\\s.exe","ExitCode":0}`
got, err := parseServices(one, `C:\WINDOWS`)
if err != nil || len(got) != 1 {
if err != nil || len(got) != 1 || got[0].State != "active" {
t.Fatalf("single object: got %+v, err %v", got, err)
}
@@ -130,6 +130,25 @@ func TestParseEventsFormatsAndOrders(t *testing.T) {
}
}
// A message containing a bare "\n" (no carriage return) must still collapse to
// one line, or it silently multiplies into several output lines and throws
// off the tail trim's count.
func TestParseEventsCollapsesBareLF(t *testing.T) {
in := `[{"t":"2026-08-13T10:00:00Z","l":"Error","p":"Contoso","m":"broker died\nstack trace here"}]`
got, err := parseEvents(in, "Contoso", "Contoso Broker", 500)
if err != nil {
t.Fatalf("parseEvents: %v", err)
}
if strings.Count(got, "\n") != 0 {
t.Fatalf("parseEvents did not collapse bare LF into one line: %q", got)
}
want := "2026-08-13T10:00:00Z Error broker died stack trace here"
if got != want {
t.Fatalf("parseEvents =\n%q\nwant\n%q", got, want)
}
}
// Service Control Manager logs every service on the host under one provider, so
// its rows must be filtered down to the target or the log is somebody else's.
func TestParseEventsFiltersOtherServicesSCM(t *testing.T) {
+1 -1
View File
@@ -41,7 +41,7 @@ export default function WorkloadsPage() {
<div className="p-4 sm:p-6 lg:p-8">
<div className="mb-6">
<h1 className="text-2xl font-bold text-text-primary">Workloads</h1>
<p className="mt-1 text-sm text-text-secondary">Containers and systemd services across the fleet, as last reported by each agent.</p>
<p className="mt-1 text-sm text-text-secondary">Containers and services across the fleet, as last reported by each agent.</p>
</div>
<Card className="mb-6">
@@ -82,7 +82,7 @@ export function MaintenanceTab({
<span className="font-mono text-xs text-text-secondary">{u.current_version || "n/a"}</span>
</Td>
<Td label={isWindows ? "KB" : "Available"}>
<span className="font-mono text-xs text-success">{u.new_version}</span>
<span className="font-mono text-xs text-success">{u.new_version || "n/a"}</span>
</Td>
</Tr>
))}
@@ -94,6 +94,7 @@ export function WorkloadList({ serverId, canControl, isWindows }: { serverId: st
workload={w}
canControl={canControl}
busy={control.isPending}
isWindows={isWindows}
onAction={(action) => control.mutate({ w, action })}
onLogs={() => setLogTarget(w)}
/>
+3 -1
View File
@@ -45,12 +45,14 @@ export function WorkloadRow({
workload,
canControl,
busy,
isWindows,
onAction,
onLogs,
}: {
workload: Workload;
canControl: boolean;
busy: boolean;
isWindows: boolean;
onAction: (action: WorkloadAction) => void;
onLogs: () => void;
}) {
@@ -66,7 +68,7 @@ export function WorkloadRow({
{!!w.restarts && w.restarts > 0 && <Badge variant="warning">{w.restarts} restarts</Badge>}
</div>
<p className="mt-1 truncate text-xs text-text-secondary">
{w.kind === "container" ? w.image || "no image" : "systemd unit"}
{w.kind === "container" ? w.image || "no image" : isWindows ? "Windows service" : "systemd unit"}
{w.ports && w.ports.length > 0 && <span className="ml-2 font-mono">{w.ports.join(" ")}</span>}
</p>
</div>
+1 -1
View File
@@ -1122,7 +1122,7 @@ export const vulnerabilities = {
export type WorkloadKind = "container" | "unit";
export type WorkloadAction = "start" | "stop" | "restart";
/** One container or one systemd unit.
/** One Docker container, one systemd unit, or one Windows service.
*
* `state` is deliberately not a shared vocabulary across the two kinds:
* containers report running/exited/paused/restarting/created, units report