lib/logstorage: sort tenantIDs before returning to the client (and before searching)

app/vlselect/logsql/logsql.go

@@ -1394,6 +1394,8 @@ func ProcessTenantIDsRequest(ctx context.Context, w http.ResponseWriter, r *http
 		httpserver.Errorf(w, r, "cannot obtain tenantIDs: %s", err)
 		return
 	}
+	// Return tenants in a deterministic (sorted) order for stable API responses.
+	logstorage.SortTenantIDs(tenants)
 
 	data, err := json.Marshal(tenants)
 	if err != nil {

app/vlstorage/netselect/netselect.go

@@ -679,6 +679,7 @@ func (s *Storage) getTenantIDs(ctx context.Context, start, end int64) ([]logstor
 		tenantIDs = append(tenantIDs, tenantID)
 	}
 
+	logstorage.SortTenantIDs(tenantIDs)
 	return tenantIDs, nil
 }
 

docs/victorialogs/CHANGELOG.md

@@ -38,6 +38,7 @@ according to the following docs:
 * BUGFIX: [cluster version](https://docs.victoriametrics.com/victorialogs/cluster/): avoid `cannot connect to storage node at ...: EOF` errors after `vlselect` or `vlinsert` was idle for more than 60 seconds. See [#1440](https://github.com/VictoriaMetrics/VictoriaLogs/issues/1440).
 * BUGFIX: [vlselect](https://docs.victoriametrics.com/victorialogs/cluster/): return `502 Bad Gateway` HTTP response code for incoming queries when one of the `vlstorage` nodes runs a VictoriaLogs version with an incompatible internal API instead of `400 Bad Request`. This is consistent with the `502 Bad Gateway` response returned when a `vlstorage` node is unavailable, and it allows building a proper failover scheme in high-availability setups. See [these docs](https://docs.victoriametrics.com/victorialogs/cluster/#high-availability).
 * BUGFIX: [multi-level cluster setup](https://docs.victoriametrics.com/victorialogs/cluster/#multi-level-cluster-setup): properly return `502 Bad Gateway` HTTP response code when a `vlselect` node queries other `vlselect` nodes and the underlying `vlstorage` is unavailable, as described at [high availability](https://docs.victoriametrics.com/victorialogs/cluster/#high-availability) docs. This allows configuring proper failover schemes to a healthy cluster.
+* BUGFIX: [`/select/tenant_ids`](https://docs.victoriametrics.com/victorialogs/querying/#querying-tenants): return tenant ids in a deterministic (sorted) order. Previously the order was random across requests. See [#1575](https://github.com/VictoriaMetrics/VictoriaLogs/pull/1575).
 
 ## [v1.51.0](https://github.com/VictoriaMetrics/VictoriaLogs/releases/tag/v1.51.0)
 

lib/logstorage/storage_search.go

@@ -236,6 +236,9 @@ func (s *Storage) runQuery(qctx *QueryContext, writeBlock writeBlockResultFunc)
 }
 
 func (s *Storage) getSearchOptions(tenantIDs []TenantID, q *Query, hiddenFieldsFilters []string) *storageSearchOptions {
+	// tenantIDs must be sorted, since block search performs binary search over them.
+	SortTenantIDs(tenantIDs)
+
 	streamIDs := q.getStreamIDs()
 	sort.Slice(streamIDs, func(i, j int) bool {
 		return streamIDs[i].less(&streamIDs[j])
@@ -752,6 +755,9 @@ func (s *Storage) getTenantIDs(ctx context.Context, start, end int64) ([]TenantI
 	for k := range uniqTenantIDs {
 		tenants = append(tenants, k)
 	}
+	// Sort tenants, since they are collected from a map with random iteration order.
+	// This provides stable output and a sorted list for callers that rely on it.
+	SortTenantIDs(tenants)
 
 	return tenants, nil
 }

lib/logstorage/storage_search_test.go

@@ -675,12 +675,8 @@ func TestStorageRunQuery(t *testing.T) {
 		if err != nil {
 			t.Fatalf("unexpected error: %s", err)
 		}
-		sort.Slice(tenantIDs, func(i, j int) bool {
-			return tenantIDs[i].less(&tenantIDs[j])
-		})
-		sort.Slice(allTenantIDs, func(i, j int) bool {
-			return allTenantIDs[i].less(&allTenantIDs[j])
-		})
+		SortTenantIDs(tenantIDs)
+		SortTenantIDs(allTenantIDs)
 		if !reflect.DeepEqual(tenantIDs, allTenantIDs) {
 			t.Fatalf("unexpected GetTenantIDs result; got: %v, want: %v", tenantIDs, allTenantIDs)
 		}

lib/logstorage/tenant_id.go

@@ -4,6 +4,7 @@ import (
 	"encoding/json"
 	"fmt"
 	"net/http"
+	"sort"
 	"strconv"
 	"strings"
 
@@ -46,6 +47,13 @@ func (tid *TenantID) less(a *TenantID) bool {
 	return tid.ProjectID < a.ProjectID
 }
 
+// SortTenantIDs sorts tenantIDs in place in ascending order.
+func SortTenantIDs(tenantIDs []TenantID) {
+	sort.Slice(tenantIDs, func(i, j int) bool {
+		return tenantIDs[i].less(&tenantIDs[j])
+	})
+}
+
 func (tid *TenantID) marshalString(dst []byte) []byte {
 	n := uint64(tid.AccountID)<<32 | uint64(tid.ProjectID)
 	dst = marshalUint64Hex(dst, n)