Files
fancy-dns/internal/config/size.go
T
developer 4f49033f1b Cover the engine, config.Size and metrics; fix the five bugs they found
internal/engine, internal/config and internal/metrics had no test file at all,
and writing them turned up five real bugs.

internal/engine (23 tests, new harness_test.go): a wired engine over a fake
upstream on loopback, a local zone, a blocklist, a cache, a query log and an
injectable clock. It pins the pipeline order end to end -- ACL deny before the
local zone, local zone before blocklist, blocklist before cache, drop sends
nothing, the concurrency and rate-limit refusals, serve-stale served with TTL 1
and refreshed once per key, truncation against a 512-byte client, and exactly
one metrics record and one query-log entry per query.

Five fixes:

- config.ParseSize ignored the unit multiplier, so 512MiB parsed as 512 bytes
  and the whole memory envelope was inert for anyone who wrote a unit.

- finalize (EDNS0 and the UDP size) ran only on the cache-hit path, so a large
  uncached answer left oversized for a 512-byte client. It now runs on the
  upstream, local-zone and blocklist paths too.

- The upstream UDP client read buffer was hard-coded to 1232 while we advertise
  dns.max_udp_size, so with a larger buffer configured every answer above 1232
  bytes was read truncated and became SERVFAIL. The pool takes the size now.

- LRU.prefetchGate was a 64-slot global cap, not the single flight its comment
  promised: a burst of stale hits launched up to 64 duplicate refreshes of one
  name. It is a per-key latch now; distinct names still refresh in parallel.

- metrics.RecordMiss and RecordStaleHit moved only the Prometheus collector, so
  the Cache page's misses and stale hits read 0 for the life of the process.

AGENTS.md gains the trap that cost the most time: this version of miekg/dns
reports a character-string over 255 bytes as zero length, so such a message
silently fails to pack, and the write failure looks like a client timeout.
2026-09-27 16:03:21 +00:00

179 lines
5.6 KiB
Go

package config
import (
"fmt"
"math"
"strconv"
"strings"
"gopkg.in/yaml.v3"
)
// Size is a byte count that unmarshals from a string such as "512MiB" or
// "2GiB". A bare integer is bytes. It exists so an operator can write a memory
// budget in the same units everywhere else, instead of a raw byte count that
// has to be converted by hand.
type Size int64
// Longest suffix first, so "MIB" is not shadowed by "B" and "M" does not eat
// the "iB" of "MiB".
var sizeUnits = []struct {
suffix string
mult int64
}{
{"KIB", 1 << 10},
{"MIB", 1 << 20},
{"GIB", 1 << 30},
{"TIB", 1 << 40},
{"KB", 1000},
{"MB", 1000 * 1000},
{"GB", 1000 * 1000 * 1000},
{"TB", 1000 * 1000 * 1000 * 1000},
{"B", 1},
{"K", 1 << 10},
{"M", 1 << 20},
{"G", 1 << 30},
{"T", 1 << 40},
}
// ParseSize reads a size with an optional unit suffix. It is strict about
// junk: "512MiBx" is an error, not 512MiB with a trailing letter.
func ParseSize(s string) (Size, error) {
raw := strings.TrimSpace(s)
if raw == "" {
return 0, fmt.Errorf("invalid size %q: empty", s)
}
upper := strings.ToUpper(raw)
for _, u := range sizeUnits {
if !strings.HasSuffix(upper, u.suffix) {
continue
}
num := strings.TrimSpace(upper[:len(upper)-len(u.suffix)])
if num == "" {
return 0, fmt.Errorf("invalid size %q: no number before the unit", s)
}
n, err := strconv.ParseFloat(num, 64)
if err != nil {
return 0, fmt.Errorf("invalid size %q", s)
}
return scaleSize(n, u.mult, s)
}
n, err := strconv.ParseFloat(raw, 64)
if err != nil {
return 0, fmt.Errorf("invalid size %q", s)
}
return scaleSize(n, 1, s)
}
// scaleSize applies the unit multiplier. A bare number has mult 1, which is
// why the multiplier is a parameter rather than a constant: it used to be
// forgotten on the suffixed path, so "512MiB" parsed as 512 *bytes* and every
// memory budget written in units was off by the unit's factor.
func scaleSize(n float64, mult int64, orig string) (Size, error) {
if n < 0 {
return 0, fmt.Errorf("invalid size %q: must not be negative", orig)
}
bytes := n * float64(mult)
if bytes >= float64(math.MaxInt64) {
return 0, fmt.Errorf("invalid size %q: out of range", orig)
}
return Size(int64(bytes)), nil
}
// UnmarshalYAML accepts both "512MiB" and a bare number.
func (s *Size) UnmarshalYAML(value *yaml.Node) error {
var str string
if err := value.Decode(&str); err != nil {
return fmt.Errorf("invalid size: %w", err)
}
v, err := ParseSize(str)
if err != nil {
return err
}
*s = v
return nil
}
// MarshalYAML writes the number back as a plain integer, so a config written
// by the API round-trips as something ParseSize still accepts.
func (s Size) MarshalYAML() (any, error) { return int64(s), nil }
// String renders the size in binary units, which is what an operator reading a
// log line wants ("512MiB" rather than "536870912").
func (s Size) String() string {
switch {
case s == 0:
return "0"
case s%(1<<30) == 0:
return fmt.Sprintf("%dGiB", s/(1<<30))
case s%(1<<20) == 0:
return fmt.Sprintf("%dMiB", s/(1<<20))
case s%(1<<10) == 0:
return fmt.Sprintf("%dKiB", s/(1<<10))
default:
return fmt.Sprintf("%dB", int64(s))
}
}
// Memory is the process-wide memory envelope.
//
// It exists because the only lever the Go runtime has before a cgroup OOM kill
// is the cgroup limit itself, and a cgroup kill is the worst possible outcome
// for a resolver: an outage, not a cache miss. GOMEMLIMIT turns the same limit
// into a soft one the GC works towards, and the cache is sized to yield before
// the GC has to start thrashing.
type Memory struct {
// Limit is the soft heap ceiling handed to runtime/debug.SetMemoryLimit.
// 0 means no ceiling.
Limit Size `yaml:"limit"`
// Reserve is subtracted from Limit before it is applied, covering what the
// Go heap does not account for: goroutine stacks, runtime metadata, and —
// importantly — the bbolt mmap for the cache snapshot and the query log,
// which lives outside the Go heap entirely. A container limit with no
// reserve is a limit the runtime will overshoot.
Reserve Size `yaml:"reserve"`
// CacheMaxBytes bounds the in-memory LRU by bytes as well as by entry
// count. 0 means "derive from Limit", which is what an operator almost
// always wants: the count budget alone does not bound the footprint,
// because a single cached ANY or TXT answer can be tens of kilobytes.
CacheMaxBytes Size `yaml:"cache_max_bytes"`
}
// DefaultReserve is the non-heap allowance applied when memory.reserve is not
// set. 25% is the usual rule of thumb for a process with a large mmapped file
// and a few thousand goroutines.
const DefaultReserve = Size(256 << 20)
// EffectiveLimit is the value to hand to debug.SetMemoryLimit, already reduced
// by the reserve. Zero means "no ceiling" (math.MaxInt64).
func (m Memory) EffectiveLimit() int64 {
if m.Limit <= 0 {
return math.MaxInt64
}
reserve := m.Reserve
if reserve == 0 {
reserve = DefaultReserve
}
limit := int64(m.Limit) - int64(reserve)
if limit < 64<<20 {
// A limit below the runtime's own floor just guarantees a death
// spiral. Report the floor rather than the literal subtraction.
return 64 << 20
}
return limit
}
// EffectiveCacheBytes is the byte budget for the LRU: the configured value, or
// 60% of the effective limit when unset. The remainder belongs to the GC, the
// query log's in-memory ring, the blocklist sets and the per-query working set.
func (m Memory) EffectiveCacheBytes() int64 {
if m.CacheMaxBytes > 0 {
return int64(m.CacheMaxBytes)
}
limit := m.EffectiveLimit()
if limit == math.MaxInt64 {
return 0 // unbounded
}
return limit * 3 / 5
}