Files
monlet/server/alembic/versions/0001_baseline.py

144 lines
5.5 KiB
Python

"""baseline schema (ADR-0004)
Revision ID: 0001_baseline
Revises:
Create Date: 2026-05-26
"""
from collections.abc import Sequence
from alembic import op
revision: str = "0001_baseline"
down_revision: str | None = None
branch_labels: str | Sequence[str] | None = None
depends_on: str | Sequence[str] | None = None
def upgrade() -> None:
op.execute(
"""
CREATE TABLE agents (
agent_id TEXT PRIMARY KEY,
hostname TEXT NOT NULL,
status TEXT NOT NULL CHECK (status IN ('alive','stale','dead')) DEFAULT 'alive',
last_seen_at TIMESTAMPTZ NOT NULL,
features JSONB NOT NULL DEFAULT jsonb_build_object('push', false, 'metrics', false),
labels JSONB NOT NULL DEFAULT '{}'::jsonb,
created_at TIMESTAMPTZ NOT NULL DEFAULT now()
);
"""
)
op.execute("CREATE INDEX agents_status_idx ON agents (status);")
op.execute("CREATE INDEX agents_last_seen_idx ON agents (last_seen_at);")
op.execute(
"""
CREATE TABLE checks (
agent_id TEXT NOT NULL REFERENCES agents(agent_id) ON DELETE CASCADE,
check_id TEXT NOT NULL,
status TEXT NOT NULL CHECK (status IN ('ok','warning','critical','unknown')),
exit_code INTEGER NOT NULL,
last_observed_at TIMESTAMPTZ NOT NULL,
last_event_id UUID NOT NULL,
incident_key TEXT NOT NULL,
summary TEXT,
PRIMARY KEY (agent_id, check_id)
);
"""
)
op.execute("CREATE INDEX checks_status_idx ON checks (status);")
# events: partitioned by RANGE(received_at). Partition key is part of PK per Postgres.
# Global idempotency by event_id is enforced via event_ingest_dedup, not this table.
op.execute(
"""
CREATE TABLE events (
event_id UUID NOT NULL,
agent_id TEXT NOT NULL,
check_id TEXT NOT NULL,
observed_at TIMESTAMPTZ NOT NULL,
received_at TIMESTAMPTZ NOT NULL DEFAULT now(),
status TEXT NOT NULL CHECK (status IN ('ok','warning','critical','unknown')),
exit_code INTEGER NOT NULL,
duration_ms INTEGER NOT NULL CHECK (duration_ms >= 0),
output TEXT,
output_truncated BOOLEAN NOT NULL DEFAULT FALSE,
notifications_enabled BOOLEAN NOT NULL DEFAULT TRUE,
incident_key TEXT NOT NULL,
PRIMARY KEY (event_id, received_at)
) PARTITION BY RANGE (received_at);
"""
)
op.execute(
"CREATE INDEX events_agent_check_observed_idx ON events (agent_id, check_id, observed_at DESC);"
)
op.execute("CREATE INDEX events_observed_at_idx ON events (observed_at DESC);")
op.execute("CREATE INDEX events_received_at_idx ON events (received_at DESC);")
# Global idempotency table: server enforces event_id uniqueness here, not on the
# partitioned events table. Pruned by retention TTL (event_dedup_retention_days).
op.execute(
"""
CREATE TABLE event_ingest_dedup (
event_id UUID PRIMARY KEY,
received_at TIMESTAMPTZ NOT NULL DEFAULT now()
);
"""
)
op.execute(
"CREATE INDEX event_ingest_dedup_received_at_idx ON event_ingest_dedup (received_at);"
)
op.execute(
"""
CREATE TABLE incidents (
id UUID PRIMARY KEY,
incident_key TEXT NOT NULL,
agent_id TEXT NOT NULL,
check_id TEXT NOT NULL,
state TEXT NOT NULL CHECK (state IN ('open','resolved')),
severity TEXT NOT NULL CHECK (severity IN ('warning','critical','unknown')),
opened_at TIMESTAMPTZ NOT NULL,
resolved_at TIMESTAMPTZ,
summary TEXT,
last_event_id UUID NOT NULL
);
"""
)
op.execute(
"CREATE UNIQUE INDEX incidents_open_key_idx ON incidents (incident_key) WHERE state = 'open';"
)
op.execute("CREATE INDEX incidents_state_opened_idx ON incidents (state, opened_at DESC);")
op.execute(
"""
CREATE TABLE notification_outbox (
id UUID PRIMARY KEY,
incident_id UUID NOT NULL REFERENCES incidents(id) ON DELETE CASCADE,
notifier TEXT NOT NULL,
event_type TEXT NOT NULL CHECK (event_type IN ('firing','resolved')),
state TEXT NOT NULL CHECK (state IN ('pending','sending','sent','retry','failed','discarded')),
attempts INTEGER NOT NULL DEFAULT 0,
next_attempt_at TIMESTAMPTZ,
last_error TEXT,
payload JSONB NOT NULL,
created_at TIMESTAMPTZ NOT NULL DEFAULT now(),
updated_at TIMESTAMPTZ NOT NULL DEFAULT now()
);
"""
)
op.execute(
"CREATE INDEX outbox_pending_idx ON notification_outbox (state, next_attempt_at) WHERE state IN ('pending','retry');"
)
def downgrade() -> None:
op.execute("DROP TABLE IF EXISTS notification_outbox CASCADE;")
op.execute("DROP TABLE IF EXISTS incidents CASCADE;")
op.execute("DROP TABLE IF EXISTS event_ingest_dedup CASCADE;")
op.execute("DROP TABLE IF EXISTS events CASCADE;")
op.execute("DROP TABLE IF EXISTS checks CASCADE;")
op.execute("DROP TABLE IF EXISTS agents CASCADE;")