Keep what a consumer gives up on until a person delivers it again or drops it

Design 25 promised a dead-letter stream that did not exist: a message a
consumer gave up on stayed only in its source, which drops it after a week,
and its condition cleared when the advisories stopped (hq issue 330, ADR 0264).
This commit is contained in:
jochen
2026-10-08 18:32:58 +02:00
parent e3ec15f707
commit 826dcb91b1
25 changed files with 1268 additions and 25 deletions
+87 -4
View File
@@ -154,8 +154,9 @@ var signalsTable = []signalRow{
}},
{Row: "S9", Signal: "bus advisories: maximum deliveries, consumer deleted; the controller's own slow " +
"consumer and refused subjects", Emitter: "bus server's advisory subjects; the controller's connection",
Trigger: "any", Bound: "any occurrence; clears after an hour without another, and a deleted consumer " +
"once it exists again or the mesh no longer expects it",
Trigger: "any", Bound: "any occurrence; clears after an hour without another, a deleted consumer " +
"once it exists again or the mesh no longer expects it, and a message given up on once DEAD_LETTERS " +
"no longer holds it (novox/hq issue 330)",
Kind: "slow-consumer, max-deliveries, refused, consumer-lost", Severity: conditions.Warning, Phase: 1,
needs: func(f *signalFacts) error { return f.advisoriesErr }, watch: watchAdvisories,
newest: func(f *signalFacts) time.Time {
@@ -485,12 +486,94 @@ func watchAdvisories(f *signalFacts) []conditions.Observation {
if a.ID == "controller" {
machine = f.host
}
out = append(out, conditions.Observation{Scope: conditions.ScopeBus, ID: a.ID, Kind: a.Kind,
Machine: machine, Severity: severity, Summary: a.Said + times, Said: a.Said})
o := conditions.Observation{Scope: conditions.ScopeBus, ID: a.ID, Kind: a.Kind, Token: a.Token,
Machine: machine, Severity: severity, Summary: a.Said + times, Said: a.Said}
if a.Kind == link.AdvisoryMaxDeliveries && a.Token == link.AdvisoryNotKept {
o.Machine = consumerMachine(a.Stream, a.Consumer)
o.Headline = clip(conditions.Capital(fmt.Sprintf("%s gave up on a message, not kept",
consumerWho(a.Stream, a.Consumer))), 60)
o.Explanation = "A listener on the bus could not handle a message, and the mesh could not keep it " +
"for you yet. It tries again every minute."
o.Resolved = "Resolved: the message is kept"
}
out = append(out, o)
}
return append(out, watchDeadLetters(f)...)
}
// watchDeadLetters says each consumer that DEAD_LETTERS holds a message for (novox/hq issue 330): open
// while it holds any, so it clears when they are delivered again or dropped, never because the server
// stopped saying it.
func watchDeadLetters(f *signalFacts) []conditions.Observation {
keys := make([]string, 0, len(f.deadLetters))
for k := range f.deadLetters {
keys = append(keys, k)
}
sort.Strings(keys)
var out []conditions.Observation
for _, key := range keys {
n := f.deadLetters[key]
stream, consumer, _ := strings.Cut(key, ".")
messages, them := "a message", "it"
if n > 1 {
messages, them = fmt.Sprintf("%d messages", n), "them"
}
who := consumerWho(stream, consumer)
out = append(out, conditions.Observation{Scope: conditions.ScopeBus, ID: key, Kind: link.AdvisoryMaxDeliveries,
Machine: consumerMachine(stream, consumer), Severity: conditions.Warning,
Summary: fmt.Sprintf("%s gave up on %s; %s kept in %s until delivered again or dropped, with why, "+
"through the controller's dead-letters verb", link.ConsumerInWords(stream, consumer), messages,
map[bool]string{true: "they are", false: "it is"}[n > 1], broker.DeadLettersStream),
Said: fmt.Sprintf("%d held for %s", n, key),
Headline: clip(conditions.Capital(fmt.Sprintf("%s could not handle %s", who, messages)), 60),
Needs: fmt.Sprintf("deliver %s again or drop %s, from the mesh MCP server.", them, them),
Explanation: conditions.Capital(fmt.Sprintf("%s was handed %s several times and gave up, so what %s "+
"asked for was not done. The mesh keeps %s until you deliver %s again or drop %s.", who, messages,
them, them, them, them)),
Resolved: "Resolved: the messages it gave up on were delivered again or dropped"})
}
return out
}
// consumerWho is a durable consumer's holder as the operator says it: a module on its machine, the
// controller, or a seat's holders.
func consumerWho(stream, consumer string) string {
switch {
case consumer == broker.ControllerName:
return "the controller"
case strings.HasPrefix(stream, "SEAT_") && strings.HasSuffix(consumer, "_worker"):
seat := strings.ToLower(strings.ReplaceAll(strings.TrimSuffix(strings.TrimPrefix(consumer, "SEAT_"), "_worker"), "_", "-"))
return "the holder of " + seat
case stream == broker.EventsStream:
if node, module, ok := strings.Cut(consumer, "_"); ok {
return module + " on " + node
}
}
return "a listener on the bus"
}
// consumerMachine is the machine a module's consumer is on; empty for the others.
func consumerMachine(stream, consumer string) string {
if stream == broker.EventsStream {
if node, _, ok := strings.Cut(consumer, "_"); ok {
return node
}
}
return ""
}
// clip is words at most n characters long, cut at a word.
func clip(s string, n int) string {
if len(s) <= n {
return s
}
cut := s[:n]
if i := strings.LastIndex(cut, " "); i > 0 {
cut = cut[:i]
}
return cut
}
func watchSelfCheck(f *signalFacts) []conditions.Observation {
every := f.selfCheck.every
if every <= 0 {