Wait out a forge not served for a moment when registering a build (novox/hq issue 486)
mesh/merge-gate pass: builds build-agent, mesh-controller → ace, g14, novox, shanks; no bus step; every machine composes with the change as it did without …
mesh/repo-check pass: its merge-check.sh passed
mesh/delivery superseded: a newer head of the same pull request

A registration asked the forge while node-tools re-served after a tool-runner
rollout, heard "nothing serves gitea.gitea_api", and failed a good build's
plan. The bus's not-served answer is now link.ErrNothingServes (and a timeout
link.ErrNoAnswer), and the registration asks again for up to 90 s on those
alone, never on a refusal; past the bound it fails saying how long it waited.
This commit is contained in:
2026-10-12 11:23:03 +02:00
parent da1473e9db
commit 47205403d3
5 changed files with 227 additions and 7 deletions
+3 -2
View File
@@ -206,8 +206,9 @@ func (b OverNATS) AskTool(ctx context.Context, module, tool string, args []byte,
if err != nil {
if errors.Is(err, nats.ErrNoResponders) {
// Said at once rather than after the whole wait: nothing is subscribed to that
// subject, which is a different fact from a tool being slow.
return nil, fmt.Errorf("nothing serves %s.%s", module, tool)
// subject, which is a different fact from a tool being slow. Said as ErrNothingServes, so an
// asker can tell a tool not served now from one that refused (novox/hq issue 486).
return nil, saidAs{fmt.Sprintf("nothing serves %s.%s", module, tool), ErrNothingServes}
}
return nil, fmt.Errorf("asking %s.%s: %w", module, tool, err)
}