fix(repository): idempotent sentinel agent creation via ON CONFLICT (M-6)

Sentinel agents (server-scanner, cloud-aws-sm, cloud-azure-kv,
cloud-gcp-sm) were created on startup with a plain INSERT whose
duplicate-key error was swallowed unconditionally. That silenced every
other DB failure too (connectivity drop, permissions change, unrelated
constraint violation) — a restart after the first boot quietly
de-fanged cloud discovery and the network scanner (CWE-662, CWE-209-
adjacent).

Shape A: add AgentRepository.CreateIfNotExists using ON CONFLICT (id)
DO NOTHING RETURNING id + sql.ErrNoRows discrimination. This keeps the
strict Create semantics (duplicate-key is an error) intact for real
agent registration and gives sentinels their own idempotent path.

- repo: CreateIfNotExists returns (created bool, err error); false,nil
  on pre-existing row; false,wrapped err on anything else.
- interface: CreateIfNotExists added to AgentRepository.
- main.go: 4 sentinel sites log Error/Info/Debug distinctly.
- mocks: service + integration mocks implement the new method.
- tests: 4 new testcontainers integration tests cover first-insert,
  idempotent second-call, concurrent 16-goroutine race (exactly one
  creator, no duplicate-key panic), and pre-cancelled context
  surfacing.

Coverage gates (go test -cover): service 67.6%/55, handler 78.6%/60,
domain 92.7%/40, middleware 80.0%/30, crypto 86.7%/85. Race/vet/
golangci-lint v2.11.4 (0 issues)/govulncheck v1.2.0 clean across all
touched packages.
This commit is contained in:
shankar0123
2026-04-17 16:32:07 +00:00
parent 80450c7180
commit 27afa4463d
6 changed files with 293 additions and 11 deletions
+41 -1
View File
@@ -70,7 +70,9 @@ func (r *AgentRepository) Get(ctx context.Context, id string) (*domain.Agent, er
return agent, nil
}
// Create stores a new agent
// Create stores a new agent. Duplicate-key errors surface to the caller —
// real-agent registration paths rely on this to detect collisions. Use
// CreateIfNotExists for sentinel/bootstrap paths where re-inserts are expected.
func (r *AgentRepository) Create(ctx context.Context, agent *domain.Agent) error {
if agent.ID == "" {
agent.ID = uuid.New().String()
@@ -92,6 +94,44 @@ func (r *AgentRepository) Create(ctx context.Context, agent *domain.Agent) error
return nil
}
// CreateIfNotExists creates an agent only if the ID doesn't already exist.
// Used for sentinel agents (server-scanner, cloud-aws-sm, cloud-azure-kv,
// cloud-gcp-sm) on first boot AND on every subsequent restart/upgrade — the
// pre-M-6 code used plain INSERT, swallowed the duplicate-key error, and so
// silently swallowed every other database failure too (CWE-662 /
// CWE-209-adjacent). ON CONFLICT (id) DO NOTHING + RETURNING id +
// sql.ErrNoRows distinguishes "row already existed" (created=false, err=nil)
// from genuine errors (connectivity, permission, constraint violations
// other than the id primary key) which still surface. Returns true if the
// row was newly inserted, false if a row with the same ID already existed.
func (r *AgentRepository) CreateIfNotExists(ctx context.Context, agent *domain.Agent) (bool, error) {
if agent.ID == "" {
agent.ID = uuid.New().String()
}
var id string
err := r.db.QueryRowContext(ctx, `
INSERT INTO agents (id, name, hostname, status, last_heartbeat_at, registered_at, api_key_hash,
os, architecture, ip_address, version)
VALUES ($1, $2, $3, $4, $5, $6, $7, $8, $9, $10, $11)
ON CONFLICT (id) DO NOTHING
RETURNING id
`, agent.ID, agent.Name, agent.Hostname, agent.Status, agent.LastHeartbeatAt,
agent.RegisteredAt, agent.APIKeyHash,
agent.OS, agent.Architecture, agent.IPAddress, agent.Version).Scan(&id)
if err != nil {
if err == sql.ErrNoRows {
// ON CONFLICT DO NOTHING — a row with this ID already existed.
return false, nil
}
return false, fmt.Errorf("failed to create agent: %w", err)
}
agent.ID = id
return true, nil
}
// Update modifies an existing agent
func (r *AgentRepository) Update(ctx context.Context, agent *domain.Agent) error {
result, err := r.db.ExecContext(ctx, `