What the report says#
This page explains each part of the report. Next: Citadel's registrations export if your apps run on Citadel.
-
Apps present and scan notes: the first line counts the apps the scan read, then the notes it printed and carried on past (
totals.scan_errorsin JSON,rollcall_scan_errorsin metrics). Today a note means a source is not configured: with no image export the PHP image column isunknown, with no PROCTOR reports every app isnone.scanprints each note as it saves the snapshot, andreport --format=json --internallists them. A note is not a failed scan; a source that refuses (a bad token) stops the scan instead. -
PHP image, declared PHP floor, Laravel: apps per version, the state from the policy (
active,security_only,eol,unknown), the date security support ends and the days left (negative once past). -
Auth package and PROCTOR: apps per auth package, and per PROCTOR result (
pass,fail,none). -
MySQL 5.6 CI job: apps by the state of PROCTOR's
mysql56-cifinding (CREED'sdatabases.mdrule says an app's migrations must run in CI against MySQL 5.6):heldwhen the report carries aheldfinding,missingwhen it carries ano-jobfinding,not_applicablewhen it carries neither and came from PROCTOR 0.15.1 or newer (the check ran and had nothing to say: no migrations), andunknownfor an app with no report, one that could not be read, or a report from PROCTOR below 0.15.1, which cannot tellheldfromnot_applicable. The dashboard shows the four counts as panels under "Apps present" (held, missing, not applicable, unknown);missingwears a red badge with the word in it while any app is missing the job, and the internal page's table has each app's state in a "MySQL 5.6 CI" column beside PROCTOR's. -
Booster hop record: apps by how their last committed Booster run ended, from the
booster-hop.jsonBooster writes at the repository root after a real upgrade run:pass,fail(a hop's gate or checks failed), andnonefor an app with no record or one that could not be read (that one also gets a scan note). In the internal view each app carries itsboosterresult, thebooster_versionthat wrote the record andbooster_at, when the run finished. A pass says the app's last hop landed, not that the app is current: read it beside the PHP image. The dashboard shows the three counts as panels after the MySQL 5.6 ones (pass, fail, no record);failwears a red badge with the word in it while any app's last hop failed, and the internal page's table has each app's result in a "Booster" column before Drift. -
Drift, counted per kind:
image_eol,image_unknown,image_below_floor,floor_allows_eol,floor_unbounded,eol_past_grace(past the policy's grace days),laravel_eol,proctor_missing,proctor_failing,override_expired. -
Progress: apps that moved since the previous scan and apps that went from an EOL image to a supported one. In the internal view each upgraded app carries the day it landed.
-
EOL burn-down: for each scan day, how many apps ran an EOL image on that day. It reads the image, not the declared floor, so until an image export is configured every app is
unknownhere even when the floor table already shows EOL versions. -
The roll per version, scan after scan: each burn-down day also carries the same apps counted three ways, so a dashboard can chart which lines the fleet is leaving and which it is joining. Counts only, never an app name:
states: per support state on that day (active,security_only,eol,unknown);supportedisactiveplussecurity_only.php_image: per PHP minor of the image, in version order withunknownlast, such as{"8.1": 3, "8.4": 9, "unknown": 2}.laravel: per Laravel major, the same way. An app with no Laravel, or one Rollcall could not read, isunknown.
The breakdown is worked out from each stored snapshot when the report is built, so snapshots written before it existed (Rollcall 0.1.9 and older) get one too, and the stored snapshot format is unchanged (
schema_version1). It reaches back as far asrollcall prunekeeps snapshots (ROLLCALL_RETENTION_DAYS). A report written by 0.1.9 or older has nostates,php_imageorlaravelon its burn-down days: read that as "no breakdown", not as zero apps. -
Forced image upgrades: with Citadel's export as the image source, how many apps have a forced base-image upgrade pending, how many are due today or earlier, and the next date (
forced_upgrades, counts only). The same counts reach metrics and ADVISOR'srollcall_fleet_summary(ADVISOR). The internal per-app list carries each app'supgrade_due. All three are zero with any other image source. The dashboard shows the three as panels under "Apps present" (apps with an upgrade pending, due today or earlier, the next date ahead); "Due today or earlier" wears a red badge with the words in it while any app's date has come, and the internal page's table has each app's date in an "Upgrade due" column beside its image state, in the same badge once the report's day has reached it ("2026-11-01, date has come").
State is never decided in code: it is derived from the dates in policy.json by CREED's reader. A version the
policy does not list is unknown.