Make live route writes cheaper so the prefix handler keeps up (closes #96)

The live route tables lose their random UUID key; the row id the
prefix-length sample draws is declared as the integer key. The prefix
tables are keyed by prefix and keep only first_seen; a batch inserts only
prefixes not stored yet. A route update sets only the columns that
change. Each kept index names the queries using it. Every write goes
through one connection kept for writing, with a 1 GiB page cache; reading
connections get 16 MiB, heap limits rise to 1.5 and 2 GiB (README
"Memory"). Batch writes go in key order. A benchmark run by hand writes
250,000 updates on 5 million routes faster than the feed's 10,000 a
second; a live run follows.

Model: opus-5-5

internal/database/models.go

@@ -35,13 +35,11 @@ type ASN struct {
 }
 
 // Prefix represents an IP prefix (CIDR block) with its IP version (4 or 6)
-// and first/last seen timestamps.
+// and the time it was first seen.
 type Prefix struct {
-	ID        uuid.UUID
 	Prefix    string
 	IPVersion int // 4 or 6
 	FirstSeen time.Time
-	LastSeen  time.Time
 }
 
 // Announcement represents a BGP announcement or withdrawal event,
@@ -71,7 +69,6 @@ type ASNPeering struct {
 //
 //nolint:tagliatelle // field names are the existing JSON API
 type LiveRoute struct {
-	ID          uuid.UUID `json:"id"`
 	Prefix      string    `json:"prefix"`
 	MaskLength  int       `json:"mask_length"`
 	IPVersion   int       `json:"ip_version"`