org.graylog2.lookup.adapters.dnslookup

Class DnsClient

java.lang.Object

org.graylog2.lookup.adapters.dnslookup.DnsClient

public class DnsClient
extends Object

Constructor Summary

Constructors

Constructor and Description
DnsClient(long queryTimeout) Creates a new DNS client with the given query timeout.
DnsClient(long queryTimeout, long requestTimeout) Creates a new DNS client with the given query and request timeout.

Method Summary

Modifier and Type	Method and Description
static boolean	allIpAddressesValid(String ipAddresses)
String	getInverseAddressFormat(String ipAddress)
static boolean	isHostName(String hostName)
static boolean	isIp4Address(String ipAddress)
static boolean	isIp6Address(String ipAddress)
static void	parseReverseLookupDomain(PtrDnsAnswer.Builder dnsAnswerBuilder, String hostname) Extract the domain name (without subdomain).
List<ADnsAnswer>	resolveIPv4AddressForHostname(String hostName, boolean includeIpVersion)
List<ADnsAnswer>	resolveIPv6AddressForHostname(String hostName, boolean includeIpVersion)
PtrDnsAnswer	reverseLookup(String ipAddress)
void	start(String dnsServerIps)
void	stop()
List<TxtDnsAnswer>	txtLookup(String hostName)

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

DnsClient

public DnsClient(long queryTimeout)

Creates a new DNS client with the given query timeout. The request timeout will be the query timeout plus the DEFAULT_REQUEST_TIMEOUT_INCREMENT.

Parameters:

queryTimeout - the query timeout

See Also:

DnsClient(long, long)

DnsClient

public DnsClient(long queryTimeout,
                 long requestTimeout)

Creates a new DNS client with the given query and request timeout. The query timeout is the maximum time to wait for a DNS response from a DNS server. The request timeout is the maximum time to wait for the DNS request completion. The request timeout should always be higher than the query timeout. Background: The request timeout is used to avoid blocking a thread by waiting for the DNS resolve future to complete. This can happen when the DnsNameResolver is closed, and the event loop is shut down while a resolver request is still running.

Parameters:

queryTimeout - the query timeout

requestTimeout - the request timeout

Method Detail

start

public void start(String dnsServerIps)

stop

public void stop()

resolveIPv4AddressForHostname

public List<ADnsAnswer> resolveIPv4AddressForHostname(String hostName,
                                                      boolean includeIpVersion)
                                               throws InterruptedException,
                                                      ExecutionException,
                                                      UnknownHostException

Throws:

InterruptedException

ExecutionException

UnknownHostException

resolveIPv6AddressForHostname

public List<ADnsAnswer> resolveIPv6AddressForHostname(String hostName,
                                                      boolean includeIpVersion)
                                               throws InterruptedException,
                                                      ExecutionException,
                                                      UnknownHostException

Throws:

InterruptedException

ExecutionException

UnknownHostException

reverseLookup

public PtrDnsAnswer reverseLookup(String ipAddress)
                           throws InterruptedException,
                                  ExecutionException

Throws:

InterruptedException

ExecutionException

parseReverseLookupDomain

public static void parseReverseLookupDomain(PtrDnsAnswer.Builder dnsAnswerBuilder,
                                            String hostname)

Extract the domain name (without subdomain). The Guava InternetDomainName implementation provides a method to correctly handle this (and handles special cases for TLDs with multiple names. eg. for lb01.store.amazon.co.uk, only amazon.co.uk would be extracted). It uses https://publicsuffix.org behind the scenes.

Some domains (eg. completely randomly defined PTR domains) are not considered to have a public suffix according to Guava. For those, the only option is to manually extract the domain with string operations. This should be a rare case.

txtLookup

public List<TxtDnsAnswer> txtLookup(String hostName)
                             throws InterruptedException,
                                    ExecutionException

Throws:

InterruptedException

ExecutionException

getInverseAddressFormat

public String getInverseAddressFormat(String ipAddress)

isHostName

public static boolean isHostName(String hostName)

isIp4Address

public static boolean isIp4Address(String ipAddress)

isIp6Address

public static boolean isIp6Address(String ipAddress)

allIpAddressesValid

public static boolean allIpAddressesValid(String ipAddresses)

Parameters:

ipAddresses - A comma-separated list of IP addresses

Returns:

true if all comma-separated IP addresses are valid "8.8.4.4, 8.8.8.8" returns true "8.8.4.4, " returns true "8.8.4.4" returns true "8.8.4.4 8.8.8.8" returns false "8.8.4.4, google.com" returns false