feat: clean up report table output and add rules

internal/inspect/index_stats/index_stats.go

@@ -19,6 +19,8 @@ var IndexStatsQuery string
 
 type Result struct {
 	Name         string
+	Table        string
+	Columns      string
 	Size         string
 	Percent_used string
 	Index_scans  int64
@@ -41,9 +43,9 @@ func Run(ctx context.Context, config pgconn.Config, fsys afero.Fs, options ...fu
 		return err
 	}
 
-	table := "|Name|Size|Percent used|Index scans|Seq scans|Unused|\n|-|-|-|-|-|-|\n"
+	table := "|Name|Table|Columns|Size|Percent used|Index scans|Seq scans|Unused|\n|-|-|-|-|-|-|-|-|\n"
 	for _, r := range result {
-		table += fmt.Sprintf("|`%s`|`%s`|`%s`|`%d`|`%d`|`%t`|\n", r.Name, r.Size, r.Percent_used, r.Index_scans, r.Seq_scans, r.Unused)
+		table += fmt.Sprintf("|`%s`|`%s`|`%s`|`%s`|`%s`|`%d`|`%d`|`%t`|\n", r.Name, r.Table, r.Columns, r.Size, r.Percent_used, r.Index_scans, r.Seq_scans, r.Unused)
 	}
 	return utils.RenderTable(table)
 }

internal/inspect/index_stats/index_stats.sql

@@ -1,13 +1,20 @@
--- Combined index statistics: size, usage percent, seq scans, and mark unused
+-- Combined index statistics: size, usage percent, seq scans, mark unused, expose table + columns
 WITH idx_sizes AS (
   SELECT
     i.indexrelid AS oid,
     FORMAT('%I.%I', n.nspname, c.relname) AS name,
+    FORMAT('%I.%I', tn.nspname, tc.relname) AS table_name,
+    (
+      SELECT STRING_AGG(pg_get_indexdef(i.indexrelid, ord::int, false), ',' ORDER BY ord)
+      FROM unnest(i.indkey::int[]) WITH ORDINALITY AS k(attnum, ord)
+    ) AS columns,
     pg_relation_size(i.indexrelid) AS index_size_bytes
   FROM pg_stat_user_indexes ui
   JOIN pg_index i ON ui.indexrelid = i.indexrelid
   JOIN pg_class c ON ui.indexrelid = c.oid
   JOIN pg_namespace n ON c.relnamespace = n.oid
+  JOIN pg_class tc ON tc.oid = i.indrelid
+  JOIN pg_namespace tn ON tn.oid = tc.relnamespace
   WHERE NOT n.nspname LIKE ANY($1)
 ),
 idx_usage AS (
@@ -37,6 +44,8 @@ usage_pct AS (
 )
 SELECT
   s.name,
+  s.table_name AS "table",
+  s.columns,
   pg_size_pretty(s.index_size_bytes) AS size,
   COALESCE(up.percent_used, 0)::text || '%' AS percent_used,
   COALESCE(u.idx_scans, 0) AS index_scans,

internal/inspect/index_stats/index_stats_test.go

@@ -30,6 +30,8 @@ func TestIndexStatsCommand(t *testing.T) {
 		conn.Query(IndexStatsQuery, reset.LikeEscapeSchema(utils.InternalSchemas)).
 			Reply("SELECT 1", Result{
 				Name:         "public.test_idx",
+				Table:        "public.test",
+				Columns:      "id",
 				Size:         "1GB",
 				Percent_used: "50%",
 				Index_scans:  5,

internal/inspect/report.go

@@ -117,6 +117,8 @@ func printSummary(ctx context.Context, outDir string) error {
 		}
 		if !match.Valid {
 			match.String = "-"
+		} else if len(match.String) > 20 {
+			match.String = fmt.Sprintf("%d matches", strings.Count(match.String, ",")+1)
 		}
 		table += fmt.Sprintf("|`%s`|`%s`|`%s`|\n", r.Name, status, match.String)
 	}

internal/inspect/templates/rules.toml

@@ -19,7 +19,13 @@ pass = "✔"
 fail = "There is at least one unused index"
 
 [[rules]]
-query = "SELECT LISTAGG(name, ',') AS match FROM `db_stats.csv` WHERE index_hit_rate < 0.94 OR table_hit_rate < 0.94"
+query = "SELECT LISTAGG(i.name, ',') AS match FROM `index_stats.csv` AS i JOIN (SELECT `table`, columns FROM `index_stats.csv` GROUP BY `table`, columns HAVING COUNT(*) > 1) AS d ON i.`table` = d.`table` AND i.columns = d.columns"
+name = "No duplicate indexes"
+pass = "✔"
+fail = "There is at least one duplicate index (same columns on the same table)"
+
+[[rules]]
+query = "SELECT 'index: ' || index_hit_rate || ', table: ' || table_hit_rate AS match FROM `db_stats.csv` WHERE index_hit_rate < 0.94 OR table_hit_rate < 0.94"
 name = "Check cache hit is within acceptable bounds"
 pass = "✔"
 fail = "There is a cache hit ratio (table or index) below 94%"
@@ -31,7 +37,7 @@ pass = "✔"
 fail = "At least one table is showing sequential scans more than 10% of total row count"
 
 [[rules]]
-query = "SELECT LISTAGG(s.tbl, ',') AS match FROM `vacuum_stats.csv` s WHERE s.expect_autovacuum = 'yes' and s.rowcount > 1000;"
+query = "SELECT LISTAGG(s.name, ',') AS match FROM `vacuum_stats.csv` s WHERE s.expect_autovacuum = 'yes' and s.rowcount > 1000;"
 name = "No large tables waiting on autovacuum"
 pass = "✔"
 fail = "At least one table is waiting on autovacuum"
@@ -41,3 +47,33 @@ query = "SELECT LISTAGG(s.name, ',') AS match FROM `vacuum_stats.csv` s WHERE s.
 name = "No tables yet to be vacuumed"
 pass = "✔"
 fail = "At least one table has never had autovacuum or vacuum run on it"
+
+[[rules]]
+query = "SELECT LISTAGG(s.name, ',') AS match FROM `vacuum_stats.csv` s WHERE FLOAT(REPLACE(s.rowcount, ',', '')) > 1000 AND FLOAT(REPLACE(s.dead_rowcount, ',', '')) > 0.2 * FLOAT(REPLACE(s.rowcount, ',', ''))"
+name = "No tables with more than 20% dead rows"
+pass = "✔"
+fail = "At least one table has more than 20% dead rows"
+
+[[rules]]
+query = "SELECT LISTAGG(slot_name, ',') AS match FROM `replication_slots.csv` WHERE active = 'f'"
+name = "No inactive replication slots"
+pass = "✔"
+fail = "There is at least one inactive replication slot"
+
+[[rules]]
+query = "SELECT LISTAGG(blocked_pid, ',') AS match FROM `blocking.csv`"
+name = "No blocked queries"
+pass = "✔"
+fail = "There is at least one query blocked on another"
+
+[[rules]]
+query = "SELECT LISTAGG(pid, ',') AS match FROM `long_running_queries.csv`"
+name = "No queries running longer than 5 minutes"
+pass = "✔"
+fail = "At least one query has been running for more than 5 minutes"
+
+[[rules]]
+query = "SELECT LISTAGG(name, ',') AS match FROM `bloat.csv` WHERE bloat > 4"
+name = "No tables or indexes with bloat ratio above 4x"
+pass = "✔"
+fail = "At least one table or index is more than 4x its expected size"

pkg/config/db.go

@@ -92,7 +92,7 @@ type (
 		Seed                seed                `toml:"seed" json:"seed"`
 		Settings            settings            `toml:"settings" json:"settings"`
 		NetworkRestrictions networkRestrictions `toml:"network_restrictions" json:"network_restrictions"`
-    SslEnforcement      *sslEnforcement     `toml:"ssl_enforcement" json:"ssl_enforcement"`
+		SslEnforcement      *sslEnforcement     `toml:"ssl_enforcement" json:"ssl_enforcement"`
 		Vault               map[string]Secret   `toml:"vault" json:"vault"`
 	}