feat(admin): end a stranded life from the portal (schema 69)
Foundation / checks (client) (push) Successful in 5s
Foundation / checks (checks) (push) Successful in 6s
Foundation / checks (deploy) (push) Successful in 6s
Foundation / checks (ecs) (push) Successful in 5s
Foundation / checks (network) (push) Successful in 4s
Foundation / checks (postgres) (push) Successful in 4s
Foundation / checks (tsan) (push) Successful in 5s
Foundation / postgres-tools (push) Successful in 5m17s
Foundation / image (push) Successful in 6m23s
Foundation / server-image (push) Successful in 8m26s
Foundation / checks (client) (push) Successful in 5s
Foundation / checks (checks) (push) Successful in 6s
Foundation / checks (deploy) (push) Successful in 6s
Foundation / checks (ecs) (push) Successful in 5s
Foundation / checks (network) (push) Successful in 4s
Foundation / checks (postgres) (push) Successful in 4s
Foundation / checks (tsan) (push) Successful in 5s
Foundation / postgres-tools (push) Successful in 5m17s
Foundation / image (push) Successful in 6m23s
Foundation / server-image (push) Successful in 8m26s
ADR-0182 section 18. Region 16 refused every birth for an account with "Existing cell is already bound to this account" although no cell in it held the account: a published reservation was never ended, and the only remedy was tools/ops/end-account-life.sql as the database owner. - Account action POST /api/accounts/<uuid>/end-life?request=N. The admin API reads the life's region from the account's lives and refuses while that region's live view lists a cell holding the account (also a bud, split offspring or organism control moved to), a region it cannot view, an incomplete view, an unpublished request or a life no longer open. - Schema 69: admin_end_account_life_v1 (admin role only) ends it at the region's latest commit and appends who and why to admin_life_endings. - The account page shows End life for each published life not ended. - A region refusing a birth because the account holds a cell (life_exists) could never record it: reject_primitive_life accepted reasons 1..3 only. The reservation stayed open and unpublished and refused every later birth. Schema 69 and the C++ wrapper accept existing_life (4). - A region with no ticking cells ignored queued admin commands (such as a spawn) until something else woke it; a queued command now keeps it ticking until applied. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
This commit is contained in:
@@ -234,6 +234,7 @@ void upgrade_across_migration_and_barrier() {
|
||||
" -v migration_sha_66=$(s 0066_tuning_revisions.sql)"
|
||||
" -v migration_sha_67=$(s 0067_admin_take_and_cull.sql)"
|
||||
" -v migration_sha_68=$(s 0068_larger_region_ticks.sql)"
|
||||
" -v migration_sha_69=$(s 0069_admin_end_life.sql)"
|
||||
" -v runtime_role=evo_runtime_test -f migrations/apply.sql"
|
||||
" > /dev/null 2>&1") != 0);
|
||||
CHECK(history(*owner_db) == "1,2,3,4,5,6,7");
|
||||
@@ -249,9 +250,10 @@ void upgrade_across_migration_and_barrier() {
|
||||
CHECK(shell("PGDATABASE=" + database +
|
||||
" EVO_PG_RUNTIME_ROLE=evo_runtime_test bash \"$EVO_SOURCE_ROOT/"
|
||||
"tools/migrate-postgres.sh\" > /dev/null") == 0);
|
||||
CHECK(history(*owner_db) == "1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,"
|
||||
"25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,"
|
||||
"47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68");
|
||||
CHECK(history(*owner_db) ==
|
||||
"1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,"
|
||||
"25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,41,42,43,44,45,46,"
|
||||
"47,48,49,50,51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69");
|
||||
CHECK(scalar(*owner_db, "SELECT (c.condeferrable AND c.condeferred AND "
|
||||
"c.confrelid = 'evolution.region_commits'::regclass)::text "
|
||||
"FROM pg_catalog.pg_constraint c WHERE c.conrelid = "
|
||||
|
||||
Reference in New Issue
Block a user