Give every module of a failed send a verdict, and excuse only the wait a build's own send brought (hq issue 318 review) #146

Merged
mesh-admin merged 3 commits from fix/318-follow-up into main 2026-10-08 13:45:02 +00:00
3 changed files with 11 additions and 4 deletions
Showing only changes of commit 7ad9dbcb5d - Show all commits
+6 -2
View File
@@ -388,6 +388,10 @@ func moduleUnhealthyObservation(module, node string, rs []inventory.ResourceHeal
// reasonWords is why a resource is unhealthy, as a person reads it.
func reasonWords(r inventory.ResourceHealth) string {
// An account waiting for a new login (ADR 0252) is said in the mesh's words, not the engine's.
if r.Kind == link.KindAccount && strings.HasPrefix(r.Reason, link.ReasonRelogin) {
return fmt.Sprintf("waits for a new login of %s, which is in the group and logged in before it was", accountOf(r))
}
switch r.Reason {
case "restarting":
return fmt.Sprintf("keeps restarting (%d restart(s) counted)", r.Restarts)
@@ -452,11 +456,11 @@ func personWait(module, machine string, rs []inventory.ResourceHealth) (string,
}
}
said := fmt.Sprintf("relogin needed on %s: %s waits for a new login of %s, which is in its group and whose "+
"running session began before it was; log out of every session and in again, or reboot", machine, module,
"login began before it was; log out of %[1]s completely and log in again, or restart it", machine, module,
strings.Join(accounts, ", "))
if len(units) > 0 {
sort.Strings(units)
said += fmt.Sprintf(" (until then %s cannot run in that session)", strings.Join(units, ", "))
said += fmt.Sprintf(" (until then %s cannot run)", strings.Join(units, ", "))
}
return said, true
}
+1 -1
View File
@@ -672,7 +672,7 @@ const FromMeshMCPServer = "from the mesh MCP server; this notification cannot do
// reloginNeeds is what an account waiting for its groups needs (ADR 0252).
func reloginNeeds(node string) string {
return fmt.Sprintf("log out of every session on %s and log in again.", node)
return fmt.Sprintf("log out of %s completely and log in again, or restart it.", node)
}
// stalledWords are the plain words of a delivery held past its bound, as mesh-delivery says it.
+4 -1
View File
@@ -94,7 +94,7 @@ func TestAModuleUnhealthyAsksForARestartInWords(t *testing.T) {
// An account waiting for a new login (ADR 0252) asks for the login, held to the plain rule.
o = moduleUnhealthyObservation("openrazer", "g14", []inventory.ResourceHealth{{Kind: "account",
Resource: "operator-in-group", Target: "jochen", Reason: "relogin needed: the account is in the group"}})
if o.Needs != "log out of every session on g14 and log in again." || len(o.Actions) != 0 {
if o.Needs != "log out of g14 completely and log in again, or restart it." || len(o.Actions) != 0 {
t.Errorf("relogin: %q %+v", o.Needs, o.Actions)
}
w := conditions.Words{Headline: o.Headline, Explanation: o.Explanation, Resolved: o.Resolved, Needs: o.Needs}
@@ -107,6 +107,9 @@ func TestAModuleUnhealthyAsksForARestartInWords(t *testing.T) {
if why, ok := conditions.PlainWords(rw, "g14"); !ok || rw.Needs == "" || len(rw.Actions) != 0 {
t.Errorf("relogin-needed: %s %+v", why, rw)
}
if strings.Contains(o.Summary, "session") || !strings.Contains(o.Summary, "waits for a new login of jochen") {
t.Errorf("relogin summary: %q", o.Summary)
}
}
// **Failed units on a machine**: "shanks's service manager is degraded: 3 failed unit(s) no module places —