A failed install says whether the package database is stale, how old it is, and what fixes it (hq issue 205) #79

2026-10-03T01:55:11Z

mesh-admin commented

2026-10-03 01:55:11 +00:00

The host's half of hq issue 205. Root cause of the unreadable failure: the existing stale-index classifier (from issue 002) read the command's stdout only, and pacman writes every failed retrieving file … 404 to stderr, which the runner folds into the error — so the control node's ten-week-old database fell through to the generic message. InstallPackage now classifies over stdout and the error text: a stale database (404s across mirrors, or a failed signature) is named with its age from the sync files' mtime and the remedy (a full upgrade by the operator; the host does not sync one package because that is a partial upgrade); a genuine mirror outage with a fresh database is named as such. Tests use the exact pacman output from the issue.

Not in here: reporting the database age as a machine fact. The host's report carries no machine facts and adding one is a protocol change on both sides; it belongs with HQ's answer to the issue's first question (who keeps the machine current).

The host's half of hq issue 205. Root cause of the unreadable failure: the existing stale-index classifier (from issue 002) read the command's stdout only, and pacman writes every `failed retrieving file … 404` to stderr, which the runner folds into the error — so the control node's ten-week-old database fell through to the generic message. `InstallPackage` now classifies over stdout and the error text: a stale database (404s across mirrors, or a failed signature) is named with its age from the sync files' mtime and the remedy (a full upgrade by the operator; the host does not sync one package because that is a partial upgrade); a genuine mirror outage with a fresh database is named as such. Tests use the exact pacman output from the issue. Not in here: reporting the database age as a machine fact. The host's report carries no machine facts and adding one is a protocol change on both sides; it belongs with HQ's answer to the issue's first question (who keeps the machine current).

mesh-admin added 1 commit 2026-10-03 01:55:11 +00:00

A failed install says whether the package database is stale, how old it is, and what fixes it (hq issue 205) c9b963f8b9

pacman writes its errors to stderr, which the runner folds into the error rather than the output;
the stale-index classifier read the output alone and never saw a single 'failed retrieving file',
so a ten-week-old database on the control node reported as a wall of 404s from the mirrors. The
classifier now reads everything pacman said, knows a failed signature as the same staleness, tells a
mirror outage with a fresh database apart from it, and names the database's date and age beside the
remedy: a full upgrade by the operator, never a one-package sync, which on this distribution is a
partial upgrade. Whose job keeping the database current is stays issue 205's question.

mesh-admin merged commit 09614d94c8 into main

2026-10-03 09:44:43 +00:00

mesh-admin deleted branch fix/issue-205

2026-10-03 09:44:44 +00:00

Sign in to join this conversation.

1 Participants

Notifications

Due Date

No due date set.

Dependencies

No dependencies set.

Reference: novox/mesh-host#79