Upgrade Command Reference
Look up the upgrade check, apply, restore and status commands, every flag they take and the exit codes they return.
- Binary
- context-platform
- Exit codes
- 0 done · 2 work to do · 3 needs a person · 1 error
All commands run as <binary> upgrade <command>. The binary is context-platform, or maestrohub-lite through the deprecated alias.
| Command | What it does |
|---|---|
upgrade check | Runs every step on copies of the stores and reports what apply would do. Changes nothing. Safe while MaestroHub runs. |
upgrade apply | With MaestroHub stopped: backs up the stores it changes, runs the upgrade, verifies it, and records it. |
upgrade restore --backup <dir> | Puts back the backup set apply wrote, all of it or none of it. Refused once 3.0 has started on the data. |
upgrade status | With MaestroHub stopped: reports whether every store is at this version, the record of past upgrades, and every backup set with its age, size, whether it holds plain-text secrets, whether it can still be restored, and the command that deletes it. While MaestroHub runs it refuses (the stores are in use). |
| Flag | Commands | Meaning |
|---|---|---|
--config <file> | all | The config file you start MaestroHub with. |
--show <part>[,<part>] | check, apply, status | Print parts of the full report before the summary: needs-a-person, problems, notes, repairs, warnings, stores, first-start, ledger, backups, or all. |
--json | all | Print the report as JSON. |
--report <file> | all | Also write the report to a file. |
--work-dir <dir> | all | Where check puts its temporary copies of the stores (and status its copy of the historian's store). Removed afterwards. |
--from-config <file> | check, apply, status | Your 2.6 config.yaml. Only the store locations in it are read, so the upgrade can find and move the stores. |
--accept-all | check, apply | Accept every item that needs a person. At a terminal, apply asks instead when you leave it out. |
--accept <id>[,<id>…] | check, apply | Accept the proposals for the listed items that need a person. |
--accept-file <file> | check, apply | Same as --accept, one ID per line. |
--legacy-tz <zone> | check, apply | The time zone 2.6 ran in, as an IANA name such as Europe/Berlin. 2.6 ran every cron schedule in that zone, whatever the schedule's time zone said; the upgrade keeps each schedule there so it fires at the same times. Default UTC, the zone of the MaestroHub container images. Give your host's zone if 2.6 ran directly on a host. |
--backup-dir <dir> | apply | Where to write the backup set, instead of <storage.dataDir>/upgrade-backups. status does not list a set written elsewhere. |
--db-snapshot-taken <id> | apply | Your confirmation, with its ID, that you took a snapshot of the PostgreSQL databases the upgrade changes. |
--copy-across-volumes | apply | Copy stores kept on another volume, checked by checksum; the originals are kept, renamed. |
--data-not-in-use | apply, restore | Your confirmation that MaestroHub is stopped, when the tool cannot tell. |
--force | apply, restore | Go ahead without asking. |
--backup <dir> | restore | The backup set to put back. |
Exit codes: 0 nothing to do or done, 2 work to do (check, status), 3 needs a person, 1 error.