Fix critical issues from follow-up security audit

A second-pass audit surfaced three severe issues missed by the previous
review, each a sibling code path of a bug class that was only partially
fixed before:

- auth: JWT session login (jwtSession.go) registered its authenticator
  even when CROSS_LOGIN_JWT_HS512_KEY was unset, leaving an empty HMAC
  key. golang-jwt verifies any HS256/HS512 signature against an empty
  key, allowing unauthenticated admin token forgery. Init() now refuses
  to register without a key, with a defense-in-depth empty-key guard in
  the keyfunc.

- repository: metric names from GraphQL ([String!]) were interpolated
  raw into json_extract(footprint, "$.<name>") SQL. SQLite parses
  double-quoted strings as literals, enabling SQL injection by any
  authenticated user. Validate metric names against ^[a-zA-Z0-9_]+$ in
  jobsMetricStatisticsHistogram and buildFloatJSONCondition.

- metricstore: cluster/host line-protocol tags flowed unvalidated into
  path.Join(RootDir, cluster, host) for checkpoint/WAL files, allowing
  arbitrary file write outside the checkpoint root via NATS
  (unauthenticated) or POST /api/write. Reject path-traversal sequences
  in DecodeLine before the tags become path components.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Entire-Checkpoint: b57246993ec1
This commit is contained in:
2026-06-04 19:07:20 +02:00
parent 6f7e262f3f
commit 6d86690c76
4 changed files with 58 additions and 7 deletions

View File

@@ -336,8 +336,18 @@ func buildTimeCondition(field string, cond *config.TimeRange, query sq.SelectBui
}
}
// validMetricName guards metric/footprint names that are interpolated into the
// json_extract() path of footprint queries. SQLite treats double-quoted strings
// as string literals, so an unescaped name (e.g. containing a `"`) would allow
// SQL injection. Legitimate metric names only use these characters.
var validMetricName = regexp.MustCompile(`^[a-zA-Z0-9_]+$`)
// buildFloatJSONCondition creates a filter on a numeric field within the footprint JSON column, using BETWEEN only if required.
func buildFloatJSONCondition(jsonField string, cond *model.FloatRange, query sq.SelectBuilder) sq.SelectBuilder {
if !validMetricName.MatchString(jsonField) {
cclog.Warnf("buildFloatJSONCondition: rejecting invalid metric name %q", jsonField)
return query.Where("0 = 1")
}
query = query.Where("JSON_VALID(footprint)")
if cond.From > 0.0 && cond.To > 0.0 {
return query.Where("JSON_EXTRACT(footprint, \"$."+jsonField+"\") BETWEEN ? AND ?", cond.From, cond.To)

View File

@@ -909,6 +909,13 @@ func (r *JobRepository) jobsMetricStatisticsHistogram(
filters []*model.JobFilter,
bins *int,
) (*model.MetricHistoPoints, error) {
// The metric name is interpolated into the json_extract() path of the SQL
// below. SQLite parses double-quoted strings as literals, so reject anything
// that is not a plain metric identifier to prevent SQL injection.
if !validMetricName.MatchString(metric) {
return nil, fmt.Errorf("invalid metric name: %q", metric)
}
// Peak value defines the upper bound for binning: values are distributed across
// bins from 0 to peak. First try to get peak from filtered cluster, otherwise
// scan all clusters to find the maximum peak value.