This session covers three distinct arcs: correctness bug fixes in the
VPP sync path and frontend reducers, new config validation, and a
large polish pass on the web frontend (tighter layout, backend kebab
dialogs, live grouped-table, live config-reload re-sync).
- encap for a VIP is now derived from the backend address family,
not the VIP's. A v6 VIP with v4 backends is programmed as IP6_GRE4
(not the buggy IP6_GRE6), matching the VPP LB plugin's
requirement that encap reflects the tunnel inner family. desiredVIP
gained an Encap field populated in desiredFromFrontend.
- ActivePoolIndex now requires at least one backend in a pool to be
BOTH in StateUp AND pb.Weight>0 before the pool counts as active.
Previously a primary pool with every backend manually zeroed would
still win over a fallback with weight=100, so fallback traffic
never materialized. New TestActivePoolIndexWeightedFailover table
pins the rule in five subcases.
- SyncLBStateVIP gained a flushAddress parameter threaded through
reconcileVIP; it forces flush=true on the setASWeight call for a
specific backend regardless of the usual 0→N heuristic. Wires up
the explicit [flush] knob the CLI exposes.
- convertFrontend already enforced that backends within one frontend
share a family. New cross-frontend pass validateVIPFamilyConsistency
rejects configs where two frontends share a VIP address but carry
backends in different families — VPP's LB plugin requires every
VIP on a prefix to have the same encap type, so such a config
would fail at lb_add_del_vip_v2 time with VNET_API_ERROR_INVALID
_ARGUMENT (-73). Catching it at config load turns a silent
runtime failure into a clear startup error.
- Two new TestValidationErrors cases pin the behavior: mismatched
families reject, same-family frontends on one VIP address allowed.
- Proto adds `bool flush = 5` to SetWeightRequest. The RPC now
drives a VIP sync immediately after mutating config (fixing the
latent "weight change only takes effect at the next 30s periodic
reconcile" gap), passing flushAddress = backend IP when req.Flush
is true.
- maglevc grows an optional [flush] token: `set frontend F pool P
backend B weight N [flush]`. Implementation uses two Run closures
(runSetFrontendPoolBackendWeight and -Flush) because the tree
walker only puts slot tokens in args — literal keywords like
`flush` advance the node but don't appear in the arg list.
- docs/user-guide.md updated with the [flush] optional and a
three-paragraph explainer of the graceful-drain vs. flush
semantics at the VPP level.
- checker.ListFrontends now sorts alphabetically to match the
existing sort in ListBackends / ListHealthChecks — RPC responses
no longer shuffle VIPs per call. cmd/frontend/client.go also
sorts defensively in refreshAll so an old maglevd build renders
alphabetically too.
- backendFromProto was returning out.Transitions[n-1] as the
LastTransition, but maglevd stores (and the proto carries)
transitions newest-first, so [n-1] was actually the oldest.
Reverse on read, which normalizes the client's Transitions slice
to oldest-first and makes [n-1] genuinely the newest. LastTransition
now points at the actual latest transition record.
- applyBackendTransition (Go and TS) derives Enabled = state!="disabled"
so the two fields stay in lockstep — closed a drift window where
a recently re-enabled backend still rendered with a stuck
[disabled] tag. The tag was later removed entirely since state
and enabled carry the same information.
- Layout tightened substantially: "FRONTENDS" panel header removed,
zippy-summary and zippy-body paddings cut, backend-table row
padding dropped to 2px, per-pool <h3> removed. Pools now live in
a single consolidated table per frontend with a dedicated "pool"
column that shows the pool name only on the first row of each
group — classic grouped-table layout, maximally dense.
- Description moved inline into the Zippy summary as muted italic
text, freeing a vertical line per frontend card.
- formatVIPAddress() helper renders IPv6 VIPs as [addr]:port and
IPv4 as addr:port, matching RFC 3986 authority syntax.
- Pools with effective_weight=0 on every backend (standby
fallbacks, fully-drained primaries) render at opacity 0.35 on
their non-actions cells; the kebab column stays at full contrast
because its menu is still fully functional on standby backends.
- Config-reload propagation: a maglevd config-reload-done log
event triggers triggerConfigResync() on the frontend side —
refreshAll() runs off the event-dispatch goroutine, then a
BrowserEvent{Type:"resync"} is published through the broker.
writeEvent emits type="resync" as a named SSE frame so the
SPA's existing addEventListener("resync") handler picks it up
and calls fetchAllState → replaceAll.
- recomputeEffectiveWeights in stores/state.ts mirrors the
server-side health.EffectiveWeights logic so the SPA keeps
pool.effective_weight correct the moment a backend transitions,
without waiting for the 30s refresh. Fixed a nasty bug where
applyBackendEffectiveWeight wrote VIP-scoped vpp-lb-sync-as-*
event weights into every frontend sharing the backend,
corrupting frontends with different per-pool configured weights.
The old log-event reducer was removed; applyConfiguredWeight is
the narrower replacement used by the kebab set-weight flow.
- applyBackendTransition calls recomputeEffectiveWeights after
state updates so pool-failover transitions (primary ⇌ fallback)
reflect instantly in the UI.
- Confirmation dialogs via a new Modal primitive
(Portal-mounted to document.body, escape/click-outside close,
click-outside debounced on mousedown so mid-row-text-selection
drags don't dismiss).
- pause/resume/enable/disable each show a Modal with a consequence
paragraph explaining what hits live traffic ("will keep existing
flows", "will flush VPP's flow table", etc.). The disable commit
button is styled btn-danger red.
- set-weight action shows a Modal with a range slider (0-100,
seeded from the current configured weight, accent-colored live
numeric readout via <output>) plus a flush checkbox and a live-
swapping note/warn paragraph describing what will happen. On
commit, the SPA also updates its local store via
applyConfiguredWeight so the operator sees the new weight
immediately without waiting for the next refresh.
- ProbeHeartbeat is now state-aware: ▶ (play) at rest for up/
down/unknown backends, ⏸ (pause) for paused, ⏹ (stop) for
disabled/removed, ❤️ (heart) during an in-flight probe.
- Drop the probe-done event listener — fast probes (<10ms)
could fire probe-done in the same render tick as probe-start
and the heart would never visibly paint. Each probe-start now
runs a fixed 400ms scale-pop animation on a timer; subsequent
probe-start events reset the timer, so fast cadences produce a
continuous heart pulse.
- Fixed wrapper box (16x14 px, overflow hidden) so the row
doesn't jiggle when the glyph swaps between the narrow ▶/⏸/⏹
text glyphs and the wider ❤️ emoji.
- Brand wordmark changed from "maglev" to "vpp-maglev" and wrapped
in an <a> linking to https://git.ipng.ch/ipng/vpp-maglev. Logo
link changed to https://ipng.ch/. Both open in a new tab with
rel="noopener".
- .gitignore fix: `frontend`, `maglevc`, `maglevd` were matching
ANY file or directory with those names anywhere in the tree,
silently ignoring cmd/frontend and friends. Anchored with
leading slashes so only repo-root build artifacts match.
719 lines
22 KiB
Go
719 lines
22 KiB
Go
// Copyright (c) 2026, Pim van Pelt <pim@ipng.ch>
|
|
|
|
package config
|
|
|
|
import (
|
|
"fmt"
|
|
"net"
|
|
"os"
|
|
"regexp"
|
|
"sort"
|
|
"strconv"
|
|
"strings"
|
|
"time"
|
|
|
|
"gopkg.in/yaml.v3"
|
|
)
|
|
|
|
// Config is the top-level parsed and validated configuration.
|
|
type Config struct {
|
|
HealthChecker HealthCheckerConfig
|
|
VPP VPPConfig
|
|
HealthChecks map[string]HealthCheck
|
|
Backends map[string]Backend
|
|
Frontends map[string]Frontend
|
|
}
|
|
|
|
// HealthCheckerConfig holds global health checker settings.
|
|
type HealthCheckerConfig struct {
|
|
TransitionHistory int
|
|
Netns string // network namespace for probes; "" = current netns
|
|
}
|
|
|
|
// VPPConfig holds VPP-related configuration.
|
|
type VPPConfig struct {
|
|
LB VPPLBConfig
|
|
}
|
|
|
|
// VPPLBConfig holds load-balancer integration settings.
|
|
type VPPLBConfig struct {
|
|
// SyncInterval is how often the full dataplane reconciliation runs,
|
|
// catching drift (e.g. manual changes to VPP). Defaults to 30s.
|
|
SyncInterval time.Duration
|
|
|
|
// IPv4SrcAddress is the source address VPP uses when encapsulating
|
|
// IPv4 traffic into GRE4 tunnels to application servers. Required
|
|
// when any frontend uses an IPv4 VIP; VPP GRE encap will fail if unset.
|
|
IPv4SrcAddress net.IP
|
|
|
|
// IPv6SrcAddress is the source address VPP uses when encapsulating
|
|
// IPv6 traffic into GRE6 tunnels. Required when any frontend uses an
|
|
// IPv6 VIP; VPP GRE encap will fail if unset.
|
|
IPv6SrcAddress net.IP
|
|
|
|
// StickyBucketsPerCore is the number of buckets (per worker thread) in
|
|
// the established-flow table. Must be a power of 2. Defaults to 65536.
|
|
StickyBucketsPerCore uint32
|
|
|
|
// FlowTimeout is the idle time after which an established flow is
|
|
// removed from the table. Must be between 1 and 120 seconds inclusive.
|
|
// Defaults to 40s.
|
|
FlowTimeout time.Duration
|
|
}
|
|
|
|
// HealthCheck describes how to probe a backend.
|
|
type HealthCheck struct {
|
|
Type string
|
|
Port uint16 // destination port; required for tcp/http/https
|
|
HTTP *HTTPParams // non-nil for type http and https
|
|
TCP *TCPParams // non-nil for type tcp
|
|
ProbeIPv4Src net.IP // source address for IPv4 probes; nil = OS picks
|
|
ProbeIPv6Src net.IP // source address for IPv6 probes; nil = OS picks
|
|
Interval time.Duration
|
|
FastInterval time.Duration // optional; used while health counter is degraded
|
|
DownInterval time.Duration // optional; used while fully down
|
|
Timeout time.Duration
|
|
Rise int // default 2
|
|
Fall int // default 3
|
|
}
|
|
|
|
// HTTPParams holds validated parameters for http/https health checks.
|
|
type HTTPParams struct {
|
|
Path string
|
|
Host string // Host header; defaults to backend IP if empty
|
|
ResponseCodeMin int // inclusive lower bound; default 200
|
|
ResponseCodeMax int // inclusive upper bound; default 200
|
|
ResponseRegexp *regexp.Regexp // nil if not configured
|
|
ServerName string // TLS SNI; falls back to Host if empty (https only)
|
|
InsecureSkipVerify bool // skip TLS certificate verification (https only)
|
|
}
|
|
|
|
// TCPParams holds validated parameters for tcp health checks.
|
|
type TCPParams struct {
|
|
SSL bool
|
|
ServerName string
|
|
InsecureSkipVerify bool
|
|
}
|
|
|
|
// Backend is a single named backend server.
|
|
type Backend struct {
|
|
Address net.IP
|
|
HealthCheck string // name reference into Config.HealthChecks; "" = no probing, assume healthy
|
|
Enabled bool // default true; false = exclude from serving entirely
|
|
}
|
|
|
|
// PoolBackend is a backend reference within a pool, with pool-local weight.
|
|
type PoolBackend struct {
|
|
Weight int // 0-100, default 100
|
|
}
|
|
|
|
// Pool is an ordered tier of backends within a frontend.
|
|
type Pool struct {
|
|
Name string
|
|
Backends map[string]PoolBackend // keyed by backend name
|
|
}
|
|
|
|
// Frontend is a single virtual IP entry.
|
|
type Frontend struct {
|
|
Description string
|
|
Address net.IP
|
|
Protocol string // "tcp", "udp", or "" (all traffic)
|
|
Port uint16 // 0 means omitted (all ports)
|
|
Pools []Pool // ordered tiers; first pool with any up backend is active
|
|
SrcIPSticky bool // when true, VPP LB uses src-IP-based hashing for this VIP
|
|
}
|
|
|
|
// ---- raw YAML types --------------------------------------------------------
|
|
|
|
type rawConfig struct {
|
|
Maglev rawMaglev `yaml:"maglev"`
|
|
}
|
|
|
|
type rawMaglev struct {
|
|
HealthChecker rawHealthCheckerCfg `yaml:"healthchecker"`
|
|
VPP rawVPPCfg `yaml:"vpp"`
|
|
HealthChecks map[string]rawHealthCheck `yaml:"healthchecks"`
|
|
Backends map[string]rawBackend `yaml:"backends"`
|
|
Frontends map[string]rawFrontend `yaml:"frontends"`
|
|
}
|
|
|
|
type rawHealthCheckerCfg struct {
|
|
TransitionHistory int `yaml:"transition-history"`
|
|
Netns string `yaml:"netns"`
|
|
}
|
|
|
|
type rawVPPCfg struct {
|
|
LB rawVPPLBCfg `yaml:"lb"`
|
|
}
|
|
|
|
type rawVPPLBCfg struct {
|
|
SyncInterval string `yaml:"sync-interval"` // Go duration; default 30s
|
|
IPv4SrcAddress string `yaml:"ipv4-src-address"`
|
|
IPv6SrcAddress string `yaml:"ipv6-src-address"`
|
|
StickyBucketsPerCore *uint32 `yaml:"sticky-buckets-per-core"` // default 65536
|
|
FlowTimeout string `yaml:"flow-timeout"` // Go duration; default 40s, [1-120]s
|
|
}
|
|
|
|
type rawHealthCheck struct {
|
|
Type string `yaml:"type"`
|
|
Port uint16 `yaml:"port"`
|
|
Params rawParams `yaml:"params"`
|
|
ProbeIPv4Src string `yaml:"probe-ipv4-src"`
|
|
ProbeIPv6Src string `yaml:"probe-ipv6-src"`
|
|
Interval string `yaml:"interval"`
|
|
FastInterval string `yaml:"fast-interval"`
|
|
DownInterval string `yaml:"down-interval"`
|
|
Timeout string `yaml:"timeout"`
|
|
Rise int `yaml:"rise"`
|
|
Fall int `yaml:"fall"`
|
|
}
|
|
|
|
type rawParams struct {
|
|
// HTTP / HTTPS
|
|
Path string `yaml:"path"`
|
|
Host string `yaml:"host"`
|
|
ResponseCode string `yaml:"response-code"`
|
|
ResponseRegexp string `yaml:"response-regexp"`
|
|
ServerName string `yaml:"server-name"`
|
|
InsecureSkipVerify bool `yaml:"insecure-skip-verify"`
|
|
// TCP
|
|
SSL bool `yaml:"ssl"`
|
|
}
|
|
|
|
type rawBackend struct {
|
|
Address string `yaml:"address"`
|
|
HealthCheck string `yaml:"healthcheck"`
|
|
Enabled *bool `yaml:"enabled"` // nil → default true
|
|
}
|
|
|
|
type rawPoolBackend struct {
|
|
Weight *int `yaml:"weight"` // nil → default 100
|
|
}
|
|
|
|
type rawPool struct {
|
|
Name string `yaml:"name"`
|
|
Backends map[string]rawPoolBackend `yaml:"backends"`
|
|
}
|
|
|
|
type rawFrontend struct {
|
|
Description string `yaml:"description"`
|
|
Address string `yaml:"address"`
|
|
Protocol string `yaml:"protocol"`
|
|
Port uint16 `yaml:"port"`
|
|
Pools []rawPool `yaml:"pools"`
|
|
SrcIPSticky bool `yaml:"src-ip-sticky"`
|
|
}
|
|
|
|
// ---- Check / Load ----------------------------------------------------------
|
|
|
|
// CheckResult holds the outcome of a config file validation. Exactly one of
|
|
// ParseError and SemanticError is non-empty when the config is invalid; both
|
|
// are empty on success.
|
|
type CheckResult struct {
|
|
ParseError string // YAML could not be read or parsed
|
|
SemanticError string // YAML parsed but semantic validation failed
|
|
}
|
|
|
|
// OK reports whether the config is valid.
|
|
func (r CheckResult) OK() bool {
|
|
return r.ParseError == "" && r.SemanticError == ""
|
|
}
|
|
|
|
// Check reads and validates the config file at path, returning the parsed
|
|
// Config (nil on failure) and a CheckResult that distinguishes YAML parse
|
|
// errors from semantic validation errors.
|
|
func Check(path string) (*Config, CheckResult) {
|
|
data, err := os.ReadFile(path)
|
|
if err != nil {
|
|
return nil, CheckResult{ParseError: fmt.Sprintf("read %q: %v", path, err)}
|
|
}
|
|
var raw rawConfig
|
|
if err := yaml.Unmarshal(data, &raw); err != nil {
|
|
return nil, CheckResult{ParseError: fmt.Sprintf("parse yaml: %v", err)}
|
|
}
|
|
cfg, err := convert(&raw.Maglev)
|
|
if err != nil {
|
|
return nil, CheckResult{SemanticError: err.Error()}
|
|
}
|
|
return cfg, CheckResult{}
|
|
}
|
|
|
|
// Load reads and validates the config file at path.
|
|
func Load(path string) (*Config, error) {
|
|
cfg, result := Check(path)
|
|
if !result.OK() {
|
|
if result.ParseError != "" {
|
|
return nil, fmt.Errorf("%s", result.ParseError)
|
|
}
|
|
return nil, fmt.Errorf("%s", result.SemanticError)
|
|
}
|
|
return cfg, nil
|
|
}
|
|
|
|
// parse unmarshals raw YAML bytes and converts them into a validated Config.
|
|
// Used by tests; production code goes through Check or Load.
|
|
func parse(data []byte) (*Config, error) {
|
|
var raw rawConfig
|
|
if err := yaml.Unmarshal(data, &raw); err != nil {
|
|
return nil, fmt.Errorf("parse yaml: %v", err)
|
|
}
|
|
return convert(&raw.Maglev)
|
|
}
|
|
|
|
func convert(r *rawMaglev) (*Config, error) {
|
|
cfg := &Config{}
|
|
|
|
// ---- healthchecker --------------------------------------------------------
|
|
cfg.HealthChecker.Netns = r.HealthChecker.Netns
|
|
cfg.HealthChecker.TransitionHistory = r.HealthChecker.TransitionHistory
|
|
if cfg.HealthChecker.TransitionHistory == 0 {
|
|
cfg.HealthChecker.TransitionHistory = 5
|
|
}
|
|
if cfg.HealthChecker.TransitionHistory < 1 {
|
|
return nil, fmt.Errorf("healthchecker.transition-history must be >= 1")
|
|
}
|
|
|
|
// ---- healthchecks ---------------------------------------------------------
|
|
cfg.HealthChecks = make(map[string]HealthCheck, len(r.HealthChecks))
|
|
for name, rh := range r.HealthChecks {
|
|
hc, err := convertHealthCheck(&rh)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("healthcheck %q: %w", name, err)
|
|
}
|
|
cfg.HealthChecks[name] = hc
|
|
}
|
|
|
|
// ---- backends -------------------------------------------------------------
|
|
cfg.Backends = make(map[string]Backend, len(r.Backends))
|
|
for name, rb := range r.Backends {
|
|
b, err := convertBackend(name, &rb, cfg.HealthChecks)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("backend %q: %w", name, err)
|
|
}
|
|
cfg.Backends[name] = b
|
|
}
|
|
|
|
// ---- frontends ------------------------------------------------------------
|
|
cfg.Frontends = make(map[string]Frontend, len(r.Frontends))
|
|
for name, rf := range r.Frontends {
|
|
fe, err := convertFrontend(name, &rf, cfg.Backends)
|
|
if err != nil {
|
|
return nil, fmt.Errorf("frontend %q: %w", name, err)
|
|
}
|
|
cfg.Frontends[name] = fe
|
|
}
|
|
|
|
// ---- cross-frontend: VIP-address family consistency -----------------------
|
|
//
|
|
// VPP's LB plugin requires every VIP sharing a given IP prefix to use
|
|
// the same encap type (GRE4 vs GRE6) — even when the VIPs sit on
|
|
// different ports. The encap is determined by the backend address
|
|
// family (see internal/vpp/lbsync.go desiredFromFrontend). So two
|
|
// frontends on the same VIP address with backends in different
|
|
// families (one IPv4 pool, one IPv6 pool) cannot both be programmed
|
|
// into VPP: the second one fails at lb_add_del_vip_v2 time with
|
|
// VNET_API_ERROR_INVALID_ARGUMENT (-73). Catching it here turns the
|
|
// silent runtime failure into a clear config-load error.
|
|
if err := validateVIPFamilyConsistency(cfg); err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
// ---- vpp ------------------------------------------------------------------
|
|
// Runs last so structural errors in healthchecks/backends/frontends are
|
|
// reported first; operators fix those, then we tell them about the VPP
|
|
// src-address requirements.
|
|
if err := convertVPP(&r.VPP, &cfg.VPP); err != nil {
|
|
return nil, err
|
|
}
|
|
|
|
return cfg, nil
|
|
}
|
|
|
|
// convertVPP parses and validates the maglev.vpp section. Missing src-address
|
|
// fields are tolerated but logged at ERROR level so operators notice that VPP
|
|
// GRE encap will fail without them.
|
|
func convertVPP(r *rawVPPCfg, cfg *VPPConfig) error {
|
|
// sync-interval: default 30s, must be > 0.
|
|
if s := r.LB.SyncInterval; s != "" {
|
|
d, err := time.ParseDuration(s)
|
|
if err != nil {
|
|
return fmt.Errorf("vpp.lb.sync-interval: %w", err)
|
|
}
|
|
if d <= 0 {
|
|
return fmt.Errorf("vpp.lb.sync-interval must be > 0")
|
|
}
|
|
cfg.LB.SyncInterval = d
|
|
} else {
|
|
cfg.LB.SyncInterval = 30 * time.Second
|
|
}
|
|
|
|
// ipv4-src-address: optional here, but warned below if missing.
|
|
if s := r.LB.IPv4SrcAddress; s != "" {
|
|
ip := net.ParseIP(s)
|
|
if ip == nil || ip.To4() == nil {
|
|
return fmt.Errorf("vpp.lb.ipv4-src-address: %q is not a valid IPv4 address", s)
|
|
}
|
|
cfg.LB.IPv4SrcAddress = ip.To4()
|
|
}
|
|
|
|
// ipv6-src-address: optional here, but warned below if missing.
|
|
if s := r.LB.IPv6SrcAddress; s != "" {
|
|
ip := net.ParseIP(s)
|
|
if ip == nil || ip.To4() != nil {
|
|
return fmt.Errorf("vpp.lb.ipv6-src-address: %q is not a valid IPv6 address", s)
|
|
}
|
|
cfg.LB.IPv6SrcAddress = ip.To16()
|
|
}
|
|
|
|
// sticky-buckets-per-core: default 65536, must be power of 2.
|
|
if p := r.LB.StickyBucketsPerCore; p != nil {
|
|
n := *p
|
|
if n == 0 || n&(n-1) != 0 {
|
|
return fmt.Errorf("vpp.lb.sticky-buckets-per-core: %d must be a power of 2", n)
|
|
}
|
|
cfg.LB.StickyBucketsPerCore = n
|
|
} else {
|
|
cfg.LB.StickyBucketsPerCore = 65536
|
|
}
|
|
|
|
// flow-timeout: default 40s, must be 1-120s inclusive and a whole number of seconds.
|
|
if s := r.LB.FlowTimeout; s != "" {
|
|
d, err := time.ParseDuration(s)
|
|
if err != nil {
|
|
return fmt.Errorf("vpp.lb.flow-timeout: %w", err)
|
|
}
|
|
if d%time.Second != 0 {
|
|
return fmt.Errorf("vpp.lb.flow-timeout: %s must be a whole number of seconds", d)
|
|
}
|
|
if d < time.Second || d > 120*time.Second {
|
|
return fmt.Errorf("vpp.lb.flow-timeout: %s out of range [1s, 120s]", d)
|
|
}
|
|
cfg.LB.FlowTimeout = d
|
|
} else {
|
|
cfg.LB.FlowTimeout = 40 * time.Second
|
|
}
|
|
|
|
// A missing src address is a hard error: VPP's GRE encap needs a source,
|
|
// and every VIP we program uses GRE. Fail the config check so the
|
|
// operator cannot start maglevd with a broken setup.
|
|
if cfg.LB.IPv4SrcAddress == nil {
|
|
return fmt.Errorf("vpp.lb.ipv4-src-address must be set; VPP GRE4 encap will fail for IPv4 VIPs")
|
|
}
|
|
if cfg.LB.IPv6SrcAddress == nil {
|
|
return fmt.Errorf("vpp.lb.ipv6-src-address must be set; VPP GRE6 encap will fail for IPv6 VIPs")
|
|
}
|
|
return nil
|
|
}
|
|
|
|
func convertHealthCheck(r *rawHealthCheck) (HealthCheck, error) {
|
|
h := HealthCheck{Type: r.Type, Port: r.Port}
|
|
|
|
switch r.Type {
|
|
case "icmp":
|
|
// ICMP does not use ports.
|
|
if r.Port != 0 {
|
|
return HealthCheck{}, fmt.Errorf("type icmp does not use a port")
|
|
}
|
|
case "tcp":
|
|
if r.Port == 0 {
|
|
return HealthCheck{}, fmt.Errorf("type tcp requires port")
|
|
}
|
|
h.TCP = &TCPParams{
|
|
SSL: r.Params.SSL,
|
|
ServerName: r.Params.ServerName,
|
|
InsecureSkipVerify: r.Params.InsecureSkipVerify,
|
|
}
|
|
case "http", "https":
|
|
if r.Port == 0 {
|
|
return HealthCheck{}, fmt.Errorf("type %s requires port", r.Type)
|
|
}
|
|
if r.Params.Path == "" {
|
|
return HealthCheck{}, fmt.Errorf("type http requires params.path")
|
|
}
|
|
min, max, err := parseCodeRange(r.Params.ResponseCode, 200)
|
|
if err != nil {
|
|
return HealthCheck{}, err
|
|
}
|
|
hp := &HTTPParams{
|
|
Path: r.Params.Path,
|
|
Host: r.Params.Host,
|
|
ResponseCodeMin: min,
|
|
ResponseCodeMax: max,
|
|
InsecureSkipVerify: r.Params.InsecureSkipVerify,
|
|
}
|
|
// TLS SNI: server-name takes precedence, falls back to host.
|
|
hp.ServerName = r.Params.ServerName
|
|
if hp.ServerName == "" {
|
|
hp.ServerName = r.Params.Host
|
|
}
|
|
if r.Params.ResponseRegexp != "" {
|
|
re, err := regexp.Compile(r.Params.ResponseRegexp)
|
|
if err != nil {
|
|
return HealthCheck{}, fmt.Errorf("invalid response-regexp %q: %w", r.Params.ResponseRegexp, err)
|
|
}
|
|
hp.ResponseRegexp = re
|
|
}
|
|
h.HTTP = hp
|
|
default:
|
|
return HealthCheck{}, fmt.Errorf("type must be \"icmp\", \"tcp\", \"http\", or \"https\", got %q", r.Type)
|
|
}
|
|
|
|
var err error
|
|
if r.ProbeIPv4Src != "" {
|
|
if h.ProbeIPv4Src, err = parseOptionalIPFamily(r.ProbeIPv4Src, 4, "probe-ipv4-src"); err != nil {
|
|
return HealthCheck{}, err
|
|
}
|
|
}
|
|
if r.ProbeIPv6Src != "" {
|
|
if h.ProbeIPv6Src, err = parseOptionalIPFamily(r.ProbeIPv6Src, 6, "probe-ipv6-src"); err != nil {
|
|
return HealthCheck{}, err
|
|
}
|
|
}
|
|
|
|
if r.Interval == "" {
|
|
return HealthCheck{}, fmt.Errorf("interval is required")
|
|
}
|
|
if h.Interval, err = time.ParseDuration(r.Interval); err != nil || h.Interval <= 0 {
|
|
return HealthCheck{}, fmt.Errorf("interval %q must be a positive duration", r.Interval)
|
|
}
|
|
if r.FastInterval != "" {
|
|
if h.FastInterval, err = time.ParseDuration(r.FastInterval); err != nil || h.FastInterval <= 0 {
|
|
return HealthCheck{}, fmt.Errorf("fast-interval %q must be a positive duration", r.FastInterval)
|
|
}
|
|
}
|
|
if r.DownInterval != "" {
|
|
if h.DownInterval, err = time.ParseDuration(r.DownInterval); err != nil || h.DownInterval <= 0 {
|
|
return HealthCheck{}, fmt.Errorf("down-interval %q must be a positive duration", r.DownInterval)
|
|
}
|
|
}
|
|
if r.Timeout == "" {
|
|
return HealthCheck{}, fmt.Errorf("timeout is required")
|
|
}
|
|
if h.Timeout, err = time.ParseDuration(r.Timeout); err != nil || h.Timeout <= 0 {
|
|
return HealthCheck{}, fmt.Errorf("timeout %q must be a positive duration", r.Timeout)
|
|
}
|
|
|
|
h.Fall = r.Fall
|
|
if h.Fall == 0 {
|
|
h.Fall = 3
|
|
}
|
|
if h.Fall < 1 {
|
|
return HealthCheck{}, fmt.Errorf("fall must be >= 1")
|
|
}
|
|
h.Rise = r.Rise
|
|
if h.Rise == 0 {
|
|
h.Rise = 2
|
|
}
|
|
if h.Rise < 1 {
|
|
return HealthCheck{}, fmt.Errorf("rise must be >= 1")
|
|
}
|
|
|
|
return h, nil
|
|
}
|
|
|
|
func convertBackend(name string, r *rawBackend, hcs map[string]HealthCheck) (Backend, error) {
|
|
ip := net.ParseIP(r.Address)
|
|
if ip == nil {
|
|
return Backend{}, fmt.Errorf("invalid address %q", r.Address)
|
|
}
|
|
|
|
b := Backend{
|
|
Address: ip,
|
|
HealthCheck: r.HealthCheck,
|
|
Enabled: boolDefault(r.Enabled, true),
|
|
}
|
|
|
|
if b.HealthCheck != "" {
|
|
if _, ok := hcs[b.HealthCheck]; !ok {
|
|
return Backend{}, fmt.Errorf("healthcheck %q not defined", b.HealthCheck)
|
|
}
|
|
}
|
|
|
|
return b, nil
|
|
}
|
|
|
|
func convertFrontend(name string, r *rawFrontend, backends map[string]Backend) (Frontend, error) {
|
|
fe := Frontend{
|
|
Description: r.Description,
|
|
Protocol: r.Protocol,
|
|
Port: r.Port,
|
|
SrcIPSticky: r.SrcIPSticky,
|
|
}
|
|
|
|
ip := net.ParseIP(r.Address)
|
|
if ip == nil {
|
|
return Frontend{}, fmt.Errorf("invalid address %q", r.Address)
|
|
}
|
|
fe.Address = ip
|
|
|
|
switch r.Protocol {
|
|
case "", "tcp", "udp":
|
|
default:
|
|
return Frontend{}, fmt.Errorf("protocol must be \"tcp\", \"udp\", or omitted, got %q", r.Protocol)
|
|
}
|
|
if r.Port != 0 && r.Protocol == "" {
|
|
return Frontend{}, fmt.Errorf("port requires protocol to be set")
|
|
}
|
|
if r.Protocol != "" && r.Port == 0 {
|
|
return Frontend{}, fmt.Errorf("protocol %q requires port to be set (1-65535)", r.Protocol)
|
|
}
|
|
|
|
if len(r.Pools) == 0 {
|
|
return Frontend{}, fmt.Errorf("pools must not be empty")
|
|
}
|
|
var firstFamily int
|
|
firstBackend := true
|
|
for pi, rp := range r.Pools {
|
|
if rp.Name == "" {
|
|
return Frontend{}, fmt.Errorf("pools[%d].name must not be empty", pi)
|
|
}
|
|
if len(rp.Backends) == 0 {
|
|
return Frontend{}, fmt.Errorf("pool %q backends must not be empty", rp.Name)
|
|
}
|
|
pool := Pool{Name: rp.Name, Backends: make(map[string]PoolBackend, len(rp.Backends))}
|
|
for bName, rpb := range rp.Backends {
|
|
b, ok := backends[bName]
|
|
if !ok {
|
|
return Frontend{}, fmt.Errorf("pool %q backend %q not defined", rp.Name, bName)
|
|
}
|
|
fam := ipFamily(b.Address)
|
|
if firstBackend {
|
|
firstFamily = fam
|
|
firstBackend = false
|
|
} else if fam != firstFamily {
|
|
return Frontend{}, fmt.Errorf("pool %q backend %q has different address family than first backend", rp.Name, bName)
|
|
}
|
|
w := intDefault(rpb.Weight, 100)
|
|
if w < 0 || w > 100 {
|
|
return Frontend{}, fmt.Errorf("pool %q backend %q weight %d out of range [0, 100]", rp.Name, bName, w)
|
|
}
|
|
pool.Backends[bName] = PoolBackend{Weight: w}
|
|
}
|
|
fe.Pools = append(fe.Pools, pool)
|
|
}
|
|
|
|
return fe, nil
|
|
}
|
|
|
|
// validateVIPFamilyConsistency walks cfg.Frontends, groups them by VIP
|
|
// address, and rejects any group whose members disagree on the backend
|
|
// address family used by their pools. See the call site in Parse for
|
|
// why this matters (VPP LB plugin limitation).
|
|
//
|
|
// Each frontend already has its own within-frontend family invariant
|
|
// (every backend in a frontend must share a family — enforced in
|
|
// convertFrontend). This check adds the cross-frontend dimension:
|
|
// frontends that happen to collide on the VIP address.
|
|
func validateVIPFamilyConsistency(cfg *Config) error {
|
|
type seen struct {
|
|
family int
|
|
frontendName string
|
|
}
|
|
byAddr := map[string]seen{}
|
|
// Sort frontend names so the "first frontend on this address"
|
|
// reported in errors is deterministic, independent of Go's
|
|
// randomized map iteration.
|
|
names := make([]string, 0, len(cfg.Frontends))
|
|
for name := range cfg.Frontends {
|
|
names = append(names, name)
|
|
}
|
|
sort.Strings(names)
|
|
for _, name := range names {
|
|
fe := cfg.Frontends[name]
|
|
fam := frontendBackendFamily(cfg, fe)
|
|
if fam == 0 {
|
|
continue // no valid backends; family is unknowable
|
|
}
|
|
addr := fe.Address.String()
|
|
if prev, ok := byAddr[addr]; ok {
|
|
if prev.family != fam {
|
|
return fmt.Errorf(
|
|
"frontend %q: VIP address %s is also used by frontend %q with IPv%d backends, "+
|
|
"but %q has IPv%d backends; VPP's LB plugin requires all VIPs sharing an "+
|
|
"address to use the same encap (backend family), so this config cannot be "+
|
|
"programmed — give the two frontends different VIP addresses",
|
|
name, addr, prev.frontendName, prev.family, name, fam)
|
|
}
|
|
continue
|
|
}
|
|
byAddr[addr] = seen{family: fam, frontendName: name}
|
|
}
|
|
return nil
|
|
}
|
|
|
|
// frontendBackendFamily returns the address family (4 or 6) of the
|
|
// first valid backend in the frontend's first pool. Returns 0 when no
|
|
// backend is resolvable — convertFrontend already enforces that all
|
|
// backends in a frontend share a family, so the first one is
|
|
// authoritative.
|
|
func frontendBackendFamily(cfg *Config, fe Frontend) int {
|
|
if len(fe.Pools) == 0 {
|
|
return 0
|
|
}
|
|
for bName := range fe.Pools[0].Backends {
|
|
if b, ok := cfg.Backends[bName]; ok && b.Address != nil {
|
|
return ipFamily(b.Address)
|
|
}
|
|
}
|
|
return 0
|
|
}
|
|
|
|
// ---- helpers ---------------------------------------------------------------
|
|
|
|
func parseOptionalIPFamily(s string, family int, field string) (net.IP, error) {
|
|
if s == "" {
|
|
return nil, nil
|
|
}
|
|
ip := net.ParseIP(s)
|
|
if ip == nil {
|
|
return nil, fmt.Errorf("%s %q is not a valid IP address", field, s)
|
|
}
|
|
if ipFamily(ip) != family {
|
|
return nil, fmt.Errorf("%s %q must be an IPv%d address", field, s, family)
|
|
}
|
|
return ip, nil
|
|
}
|
|
|
|
func ipFamily(ip net.IP) int {
|
|
if ip.To4() != nil {
|
|
return 4
|
|
}
|
|
return 6
|
|
}
|
|
|
|
func parseCodeRange(s string, defaultCode int) (min, max int, err error) {
|
|
if s == "" {
|
|
return defaultCode, defaultCode, nil
|
|
}
|
|
if idx := strings.IndexByte(s, '-'); idx > 0 {
|
|
min, err = strconv.Atoi(s[:idx])
|
|
if err != nil {
|
|
return 0, 0, fmt.Errorf("invalid response-code range %q", s)
|
|
}
|
|
max, err = strconv.Atoi(s[idx+1:])
|
|
if err != nil {
|
|
return 0, 0, fmt.Errorf("invalid response-code range %q", s)
|
|
}
|
|
return min, max, nil
|
|
}
|
|
min, err = strconv.Atoi(s)
|
|
if err != nil {
|
|
return 0, 0, fmt.Errorf("invalid response-code %q", s)
|
|
}
|
|
return min, min, nil
|
|
}
|
|
|
|
func boolDefault(p *bool, def bool) bool {
|
|
if p == nil {
|
|
return def
|
|
}
|
|
return *p
|
|
}
|
|
|
|
func intDefault(p *int, def int) int {
|
|
if p == nil {
|
|
return def
|
|
}
|
|
return *p
|
|
}
|