Enhance DNS failover reliability (#1637)

* Fix using wrong array index in log to avoid potential panic

* Increase gRPC connection timeout and add the timeout resolv.conf option

This makes sure the dns client is able to failover to a second
configured nameserver, if present. That is the case then when using the
dns `file` manager and a resolv.conf file generated for netbird.

* On file backup restore, remove the first NS if it's the netbird NS

* Bump dns mangager discovery message from debug to info to ease debugging
This commit is contained in:
Viktor Liu
2024-03-01 15:17:35 +01:00
committed by GitHub
parent a4b9e93217
commit 17b1099032
7 changed files with 224 additions and 15 deletions

View File

@ -5,6 +5,7 @@ package dns
import (
"fmt"
"os"
"regexp"
"strings"
log "github.com/sirupsen/logrus"
@ -14,6 +15,9 @@ const (
defaultResolvConfPath = "/etc/resolv.conf"
)
var timeoutRegex = regexp.MustCompile(`timeout:\d+`)
var attemptsRegex = regexp.MustCompile(`attempts:\d+`)
type resolvConf struct {
nameServers []string
searchDomains []string
@ -103,3 +107,62 @@ func parseResolvConfFile(resolvConfFile string) (*resolvConf, error) {
}
return rconf, nil
}
// prepareOptionsWithTimeout appends timeout to existing options if it doesn't exist,
// otherwise it adds a new option with timeout and attempts.
func prepareOptionsWithTimeout(input []string, timeout int, attempts int) []string {
configs := make([]string, len(input))
copy(configs, input)
for i, config := range configs {
if strings.HasPrefix(config, "options") {
config = strings.ReplaceAll(config, "rotate", "")
config = strings.Join(strings.Fields(config), " ")
if strings.Contains(config, "timeout:") {
config = timeoutRegex.ReplaceAllString(config, fmt.Sprintf("timeout:%d", timeout))
} else {
config = strings.Replace(config, "options ", fmt.Sprintf("options timeout:%d ", timeout), 1)
}
if strings.Contains(config, "attempts:") {
config = attemptsRegex.ReplaceAllString(config, fmt.Sprintf("attempts:%d", attempts))
} else {
config = strings.Replace(config, "options ", fmt.Sprintf("options attempts:%d ", attempts), 1)
}
configs[i] = config
return configs
}
}
return append(configs, fmt.Sprintf("options timeout:%d attempts:%d", timeout, attempts))
}
// removeFirstNbNameserver removes the given nameserver from the given file if it is in the first position
// and writes the file back to the original location
func removeFirstNbNameserver(filename, nameserverIP string) error {
resolvConf, err := parseResolvConfFile(filename)
if err != nil {
return fmt.Errorf("parse backup resolv.conf: %w", err)
}
content, err := os.ReadFile(filename)
if err != nil {
return fmt.Errorf("read %s: %w", filename, err)
}
if len(resolvConf.nameServers) > 1 && resolvConf.nameServers[0] == nameserverIP {
newContent := strings.Replace(string(content), fmt.Sprintf("nameserver %s\n", nameserverIP), "", 1)
stat, err := os.Stat(filename)
if err != nil {
return fmt.Errorf("stat %s: %w", filename, err)
}
if err := os.WriteFile(filename, []byte(newContent), stat.Mode()); err != nil {
return fmt.Errorf("write %s: %w", filename, err)
}
}
return nil
}