A seed was consulted only for a resource kind the store did not hold, so a document that changed in the build never reached a database an earlier release had seeded. For workflow-definitions that is not a stale setting but a total outage: the collection validates as one document, so a stored step type this build no longer has fails every task's pin, whatever workflow the task names. The seed's request ID was stable per kind and the store answers a known request ID from resource_history, which outlives the resource, so the documented recovery -- remove the resource and restart -- wrote nothing either. Seeds now carry one identity the reader can name, and a kind the store holds is refreshed only when the seed itself wrote the newest revision: a file-derived value stays the store's, and an operator's replacement is never undone. The request ID is derived from the value, so an idempotent replay is still one revision without making a second, changed document impossible. A replay whose resource is gone re-creates it from the ledger's revision. The shipped-vs-file-derived distinction is the seed's to make, so it is a property of the definition (Defaults is set exactly for the kinds that ship a document) rather than a check at each consumer.
57 lines
2.1 KiB
SQL
57 lines
2.1 KiB
SQL
-- name: LockResourceWrite :exec
|
|
-- This lock is acquired in its own statement before a resource read. PostgreSQL
|
|
-- takes a fresh Read Committed snapshot for the subsequent read, so concurrent
|
|
-- optimistic writers see the committed revision rather than both writing v1.
|
|
SELECT pg_advisory_xact_lock(hashtextextended($1, 0));
|
|
|
|
-- name: ResourceByKind :one
|
|
SELECT kind, value, version, updated_at, org_id FROM resources WHERE kind = $1;
|
|
|
|
-- name: ResourceByRequest :one
|
|
SELECT kind, value, version, actor, source, request_id, expected_version,
|
|
current_version, at
|
|
FROM resource_history
|
|
WHERE kind = $1 AND request_id = $2;
|
|
|
|
-- name: ReinsertResource :one
|
|
-- Re-creates the current row for a revision resource_history already records.
|
|
-- The ledger outlives the row it describes, so a resource an operator removed
|
|
-- still has its revisions; putting one back needs the row re-created without a
|
|
-- second ledger entry, which idx_resource_history_kind_request would refuse.
|
|
-- The revision is the ledger's, not a fresh one: this is the row that write
|
|
-- produced, restored.
|
|
INSERT INTO resources (kind, value, version, updated_at)
|
|
VALUES ($1, $2, $3, $4)
|
|
RETURNING kind, value, version, updated_at, org_id;
|
|
|
|
-- name: ResourceHistory :many
|
|
SELECT kind, value, version, actor, source, request_id, expected_version,
|
|
current_version, at
|
|
FROM resource_history
|
|
WHERE kind = $1
|
|
ORDER BY version DESC
|
|
LIMIT NULLIF(sqlc.arg(entry_limit)::bigint, 0);
|
|
|
|
-- name: ResourceVersion :one
|
|
SELECT version FROM resources WHERE kind = $1;
|
|
|
|
-- name: InsertResource :one
|
|
INSERT INTO resources (kind, value, version, updated_at)
|
|
VALUES ($1, $2, 1, $3)
|
|
RETURNING kind, value, version, updated_at, org_id;
|
|
|
|
-- name: UpdateResource :one
|
|
UPDATE resources
|
|
SET value = $2, version = version + 1, updated_at = $3
|
|
WHERE kind = $1 AND version = $4
|
|
RETURNING kind, value, version, updated_at, org_id;
|
|
|
|
-- name: InsertResourceHistory :one
|
|
INSERT INTO resource_history (
|
|
kind, value, version, actor, source, request_id, expected_version,
|
|
current_version, at
|
|
)
|
|
VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9)
|
|
RETURNING kind, value, version, actor, source, request_id, expected_version,
|
|
current_version, at;
|