Adds a self-hosted (Ollama, qwen2.5-coder) model provider abstraction with a pluggable opt-in cloud adapter, schema grounding, and a shared cost/safety guard every AI-suggested query is assessed against -- compiling to and executing through the same unchanged Phase 2 IR/ compiler and Phase 4 tenant scoping as a hand-written query, no parallel execution path. Track A (built into the query bar): inline ghost-text autocomplete, "Explain this query", "Fix this query" with a diff view, and a rule-based "Optimize" suggestion. Track B: natural-language-to-query translation, always a separate review step from execution, with `sentryctl query --nl` requiring explicit confirmation to run. Every accepted/dismissed translate-fix-optimize interaction is logged into the same append-only audit_log table Phase 4 built. Two real product bugs were found and fixed via live browser verification (a Svelte effect re-running on every keystroke that silently cancelled the ghost-text debounce; a ghost-text widget positioned at document offset 0 instead of the cursor), and a real costguard logic bug (unbounded-aggregation vs. raw-row) was caught by its own test suite. New integration tests wire a real Ollama client through the real HTTP handler against a mock server matching Ollama's wire contract (hack/mock-ollama), keeping model-quality verification out of CI as a disclosed, periodic human-run check instead. See /docs/phase-7-ai-design.md and /docs/phase-7-runbook.md.
206 lines
7.8 KiB
Go
206 lines
7.8 KiB
Go
// Package queryapi is Sentry's query API: a single POST /query endpoint
|
|
// accepting either the pipe syntax or raw SQL, compiled by
|
|
// querylang/planner and executed by querylang/executor. Replaces Phase
|
|
// 0/1's two separate placeholder endpoints (raw-SQL-only /query,
|
|
// free-text-only /search) -- see /docs/query-language-design.md.
|
|
//
|
|
// Still plain net/http, not the pinned gRPC+REST-gateway pattern, for
|
|
// the same reason as Phase 0/1: this is one endpoint, and the
|
|
// proto/annotations/codegen machinery doesn't buy much at that size.
|
|
// `/api` does speak gRPC internally (to /search) — this simplification
|
|
// is about the public-facing surface only.
|
|
package queryapi
|
|
|
|
import (
|
|
"context"
|
|
"encoding/json"
|
|
"log/slog"
|
|
"net/http"
|
|
"strings"
|
|
"time"
|
|
|
|
"github.com/sentry/sentry/api/ai/costguard"
|
|
"github.com/sentry/sentry/api/authz"
|
|
"github.com/sentry/sentry/api/internal/querylang/planner"
|
|
"github.com/sentry/sentry/api/querylang/executor"
|
|
)
|
|
|
|
// AuditLogger is core's extension point for query audit logging --
|
|
// deliberately minimal and tenant-agnostic, since core has no concept of
|
|
// tenants (see /docs/phase-4-isolation-design.md: that mechanism lives
|
|
// entirely in enterprise/). enterprise/internal/audit implements this
|
|
// against the real hash-chained, append-only store; a nil AuditLogger
|
|
// (the default for a single-tenant deployment without enterprise/
|
|
// configured) means no audit logging happens and core behaves exactly
|
|
// as it did in Phases 0-3.
|
|
//
|
|
// Tenant/user identity is deliberately NOT a field on QueryAuditEntry --
|
|
// once Phase 4 task 5's auth middleware wraps this handler, it attaches
|
|
// that identity to the request's context.Context via
|
|
// enterprise/internal/tenant, and LogQuery's ctx parameter is the same
|
|
// context the request carried, so an enterprise-side implementation
|
|
// reads identity from ctx rather than this interface growing
|
|
// tenant-awareness. Per /docs/phase-4-isolation-design.md's audit
|
|
// section, this is a fail-open path: a LogQuery error is logged but
|
|
// never fails the HTTP response for a routine read query.
|
|
type AuditLogger interface {
|
|
LogQuery(ctx context.Context, entry QueryAuditEntry) error
|
|
}
|
|
|
|
type QueryAuditEntry struct {
|
|
Query string
|
|
Language string
|
|
RowCount int
|
|
Duration time.Duration
|
|
Success bool
|
|
Error string
|
|
}
|
|
|
|
type Handler struct {
|
|
logger *slog.Logger
|
|
sqlRunner executor.SQLRunner
|
|
search executor.SearchClient
|
|
queryTimeout time.Duration
|
|
audit AuditLogger
|
|
authorizer authz.Authorizer
|
|
}
|
|
|
|
// audit and authorizer may both be nil -- see AuditLogger's doc comment
|
|
// and authz.RequireRoleOrService's nil-safety. /query allows RoleViewer
|
|
// (human sessions) or the alerting service identity (RoleService) --
|
|
// it's the one endpoint /alerting's evaluator legitimately calls, per
|
|
// /docs/phase-4-isolation-design.md's alerting service-identity design.
|
|
func NewHandler(logger *slog.Logger, sqlRunner executor.SQLRunner, search executor.SearchClient, queryTimeout time.Duration, audit AuditLogger, authorizer authz.Authorizer) *Handler {
|
|
return &Handler{logger: logger, sqlRunner: sqlRunner, search: search, queryTimeout: queryTimeout, audit: audit, authorizer: authorizer}
|
|
}
|
|
|
|
// RegisterRoutes adds this handler's routes onto a shared mux. Phase 3
|
|
// introduced a second handler package (dashboards), so CORS is
|
|
// now applied once, by main.go, around the fully-assembled mux rather
|
|
// than by each handler wrapping itself individually -- see
|
|
// httpserver.WithCORS.
|
|
func (h *Handler) RegisterRoutes(mux *http.ServeMux) {
|
|
mux.HandleFunc("POST /query", authz.RequireRoleOrService(h.authorizer, authz.RoleViewer, h.handleQuery))
|
|
mux.HandleFunc("GET /healthz", h.handleHealthz)
|
|
}
|
|
|
|
func (h *Handler) handleHealthz(w http.ResponseWriter, _ *http.Request) {
|
|
w.WriteHeader(http.StatusOK)
|
|
}
|
|
|
|
type queryRequest struct {
|
|
Query string `json:"query"`
|
|
// Language overrides auto-detection ("" / omitted). "sql" or "spl" --
|
|
// see planner.Language and /docs/query-language-design.md's
|
|
// "Detection" section for why this exists: the rare case a pipe
|
|
// query legitimately starts with the literal word "select".
|
|
Language string `json:"language"`
|
|
}
|
|
|
|
type queryResponse struct {
|
|
Columns []string `json:"columns"`
|
|
Rows [][]any `json:"rows"`
|
|
// Warnings surfaces costguard's assessment (Phase 7 task 4) for
|
|
// every query, hand-written or AI-suggested alike -- the same
|
|
// guard, never a hard block here. AI-suggested queries get a
|
|
// stricter treatment (a Reject-level assessment withholds the
|
|
// suggestion entirely, see the ai package) before a query ever
|
|
// reaches this handler; a hand-written query submitted directly
|
|
// always runs regardless of what this says, matching every prior
|
|
// phase's behavior -- this field is informational, not new
|
|
// enforcement, a deliberate choice recorded in
|
|
// /docs/phase-7-ai-design.md rather than a retrofit nobody decided
|
|
// on. Omitted (not an empty array) when there's nothing to say, so
|
|
// existing callers that don't look for this field see no shape
|
|
// change at all.
|
|
Warnings []string `json:"warnings,omitempty"`
|
|
}
|
|
|
|
type errorResponse struct {
|
|
Error string `json:"error"`
|
|
}
|
|
|
|
// maxBodyBytes caps the request body: a query string has no legitimate
|
|
// reason to be larger than this.
|
|
const maxBodyBytes = 1 << 20 // 1 MiB
|
|
|
|
func (h *Handler) handleQuery(w http.ResponseWriter, r *http.Request) {
|
|
r.Body = http.MaxBytesReader(w, r.Body, maxBodyBytes)
|
|
|
|
var req queryRequest
|
|
if err := json.NewDecoder(r.Body).Decode(&req); err != nil {
|
|
writeError(w, http.StatusBadRequest, "invalid JSON body: "+err.Error())
|
|
return
|
|
}
|
|
if strings.TrimSpace(req.Query) == "" {
|
|
writeError(w, http.StatusBadRequest, "query must not be empty")
|
|
return
|
|
}
|
|
|
|
lang := planner.Language(req.Language)
|
|
if lang != planner.Auto && lang != planner.SQL && lang != planner.SPL {
|
|
writeError(w, http.StatusBadRequest, `language must be "sql", "spl", or omitted`)
|
|
return
|
|
}
|
|
|
|
plan, err := planner.Compile(req.Query, lang, time.Now())
|
|
if err != nil {
|
|
writeError(w, http.StatusBadRequest, err.Error())
|
|
return
|
|
}
|
|
|
|
ctx, cancel := context.WithTimeout(r.Context(), h.queryTimeout)
|
|
defer cancel()
|
|
|
|
start := time.Now()
|
|
result, err := executor.Execute(ctx, plan, h.sqlRunner, h.search)
|
|
duration := time.Since(start)
|
|
|
|
if err != nil {
|
|
h.logger.Error("query execution failed", "query", req.Query, "error", err)
|
|
h.logAudit(r.Context(), req, 0, duration, err)
|
|
writeError(w, http.StatusBadGateway, "query failed: "+err.Error())
|
|
return
|
|
}
|
|
|
|
h.logAudit(r.Context(), req, len(result.Rows), duration, nil)
|
|
resp := queryResponse{Columns: result.Columns, Rows: result.Rows}
|
|
if assessment := costguard.Assess(plan); assessment.Level != costguard.LevelOK {
|
|
resp.Warnings = []string{costguard.Summary(assessment)}
|
|
}
|
|
writeJSON(w, resp)
|
|
}
|
|
|
|
// logAudit is fail-open by design (see AuditLogger's doc comment): a
|
|
// write failure here is logged and otherwise ignored, never surfaced to
|
|
// the HTTP caller. Uses r.Context() (the original request context, not
|
|
// the query-execution one with its own deadline) so a slow/cancelled
|
|
// query's context.WithTimeout expiring doesn't also cancel the audit
|
|
// write for it.
|
|
func (h *Handler) logAudit(ctx context.Context, req queryRequest, rowCount int, duration time.Duration, execErr error) {
|
|
if h.audit == nil {
|
|
return
|
|
}
|
|
entry := QueryAuditEntry{
|
|
Query: req.Query, Language: req.Language, RowCount: rowCount,
|
|
Duration: duration, Success: execErr == nil,
|
|
}
|
|
if execErr != nil {
|
|
entry.Error = execErr.Error()
|
|
}
|
|
if err := h.audit.LogQuery(ctx, entry); err != nil {
|
|
h.logger.Error("audit log write failed", "error", err)
|
|
}
|
|
}
|
|
|
|
func writeJSON(w http.ResponseWriter, v any) {
|
|
w.Header().Set("Content-Type", "application/json")
|
|
_ = json.NewEncoder(w).Encode(v)
|
|
}
|
|
|
|
func writeError(w http.ResponseWriter, status int, msg string) {
|
|
w.Header().Set("Content-Type", "application/json")
|
|
w.WriteHeader(status)
|
|
_ = json.NewEncoder(w).Encode(errorResponse{Error: msg})
|
|
}
|