The DNS subsystem in SER can either directly use
libresolv and a combination of the locally configured DNS server,
/etc/hosts and the local Network Information Service (NIS/YP and so on) or cache the query results (both positive and negative) and look first in its internal cache.
When its internal DNS cache is enabled, SER can also use DNS failover: if one destination resolves to multiple addresses SER can try all of them until it finds one to which it can successfully send the packet or it exhausts all of them. SER (tm module to be more precise) uses the DNS failover also when the destination host doesn’t send any reply to a forwarded INVITE within the SIP timeout interval (whose value can be configured using the tm fr_timer parameter). When SRV based load balancing is enabled SER can even do DNS based load balancing (see RFC 2782 and the dns_srv_lb option below).
Using the DNS cache and the DNS failover has also some drawbacks:
/etc/resolv.conf) is used for the requests (
/etc/hostsand the local Network Information Service are ignored).
use_dns_cache=offor compile without
-DUSE_DNS_FAILOVER). Turning it off from the config file is not enough in this case (the extra memory will still be used).
On the other hand using the DNS cache saves lots of DNS queries and makes DNS based failover and DNS based load balancing possible. If the destination blacklist is enabled, SER can do failover even if forwarding in stateless mode.
In the ideal case with DNS cache enabled SER will do only one query for a NAPTR (if enabled) or SRV lookup and then it will use the results for the record’s TTL (for example if all the resulting records have 1 minute TTL, SER won’t make another query for this domain for 1 minute). Even negative answers will be cached.
Without the DNS cache, each NAPTR or SRV lookup will result in at least 2 queries. These queries will happen every time, for each message (even if all of them go to the same domain).
These are the options that can be set in the configuration file of SER and control the behavior of the DNS subsystem.
When SER receives a SIP request, it compares the source IP address of the IP datagram with the IP address or hostname in the topmost
Via header field. If they are different then SER will add
received parameter to the topmost
Via header field.
Via header fields can contain both IP addresses and hostnames.
If this parameter is set to 1 and the topmost Via header field contains a hostname then SER will resolve the hostname and then it will compare the source IP address of the datagram with all the IP addresses received from DNS.
If you set the parameter to 0 then SER will not match the source IP address with the hostname in topmost Via and all such Via header fields will be modified to include the
received parameter containing the source IP.
This parameter is set to 0 by default, because vast majority of SIP implementations put IP addresses into
Via headers and thus no hostname resolution is necessary.
You should only turn this feature on if you have any user agents that put hostnames in Via headers and if you want to make sure that SIP replies will be routed to the IP address associated with the hostname in topmost Via and not to the source IP address of the SIP request.
This option is just another name for dns option described above.
This parameter controls whether SER performs reverse DNS lookups to translate the source IP address of the datagram to hostname when comparing the source IP address with the address or hostname in the topmost Via header field.
This parameter is turned off by default (SER does not perform reverse DNS lookups when comparing the the hostname in Via).
The DNS resolver options control how SER will interact with the external DNS servers. These options (with the
dns_try_ipv6 exception) are passed to
libresolv and are used each time a DNS request is made.
The default values are system specific and generally depend on the
/etc/resolv.conf content. For servers doing a lot of DNS requests it is highly recommended to change the default values in the SER config file (even if using SER’s internal DNS cache).
dns_try_ipv6 = on | off
If on and SER listens on at least one IPv6 socket, IPv6 (AAAA) lookups will be performed if the IPv4 (A) lookups fail. If off only IPv4 (A) lookups will be used.
Default: on if SER is compiled with IPv6 support.
dns_try_naptr = on | off
If on SER will first try a NAPTR lookup for destinations that don’t have the protocol or port specified and are not simple IP addresses (as described in RFC 3263). This will introduce a slight performance penalty and will probably cause extra DNS lookups. For example a lookup for a non-existing domain will produce one extra query: NAPTR(domain), SRV(_sip._udp.domain) and A/AAAA(domain).
If the result of a query contains several NAPTR records, ser will select among them according to the RFC 2915 and ser preference towards a specific protocol (see dns_udp_pref, dns_tcp_pref and dns_tls_pref below). For an RFC 3263 compliant configuration (choose the remote side preferred protocol if supported), set dns_udp_pref, dns_tcp_pref and dns_tls_pref to the same value (>=0), e.g. 0.
dns_udp_pref = number
UDP protocol preference when doing NAPTR lookups. This option works together with dns_tcp_pref and dns_tls_pref. If all this options have the same positive value and more NAPTR records are available, SER will select the NAPTR record preferred by the remote side (according to RFC 2915). If the values are positive but different, SER will select the NAPTR record whose protocol it prefers the most (the protocol with the highest dns_<proto>_pref number). If there are several NAPTR records with the same preferred protocol, SER will select among them based on their order and preference (see RFC 2915).
To completely disable selecting a specific protocol, use a negative number. For example dns_tcp_pref=-1 will completely disable selection of TCP NAPTR records, even if this will result in the NAPTR lookup failure.
Default: dns_udp_pref=3, dns_tcp_pref=2 and dns_tls_pref=1 (prefer UDP, but if no UDP NAPTR record found or no SRV-resolvable UDP NAPTR record found use TCP records and if this fails too use TLS)
dns_tcp_pref = number(see dns_udp_pref above)
dns_tls_pref = number
(see dns_udp_pref above)
dns_retr_time = timeTime in seconds before retrying a DNS request.
dns_retr_no = no.Number of DNS retransmissions before giving up.
dns_servers_no = no.How many DNS servers from the ones defined in /etc/resolv.conf will be used.
dns_use_search_list= yes/noIf no, the search list in /etc/resolv.conf will be ignored (⇒ fewer lookups ⇒ gives up faster).
HINT: even if you don’t have a search list defined, setting this option to “no” will still be “faster”, because an empty search list is in fact search “” (so even if the search list is empty/missing there will still be 2 DNS queries, eg. foo+’.’ and foo+”“+’.’)
The maximum time a dns request can take (before failing) is: (dns_retr_time*dns_retr_no)*(search_list_domains) If dns_try_ipv6 is yes, mutliply it again by 2.
The option combination that produces the “fastest” dns resolver config (the “faster” in the sense that it gives up the quickest) is:
dns_try_ipv6=no dns_retr_time=1 dns_retr_no=1 dns_servers_no=1 dns_use_search_list=no
The recommended DNS configuration is to have a “close” DNS caching recursive server configured in /etc/resolv.conf, set the DNS resolver options in SER’s configuration file as in the above example and enable the DNS cache (in SER). Pay particular attention to dns_servers_no and dns_use_search_list. It’s a good idea to make sure you don’t need / use the search list or more then one DNS server (to avoid unnecessary extra lookups).
If defined the naptr lookup support will be compiled in. NAPTR support still has to be enabled from SER’s config file (it’s off by default).
If defined, the resolver will be very verbose: it will log a lot of debugging information at L_DBG level.
If defined the NAPTR related resolver functions will be very verbose.
use_dns_cache = on | offIf off the DNS cache won’t be used (all DNS lookups will result into a DNS request). When on all the DNS request results will be cached.
use_dns_failover = on |offIf on and sending a request fails (due to not being allowed from an onsend_route, send failure, blacklisted destination or, when using tm, invite timeout), and the destination resolves to multiple IP addresses and/or multiple SRV records, the send will be re-tried using the next IP/record. In tm’s case a new branch will be created for each new send attempt.
dns_srv_lb = on | off or dns_srv_loadbalancing = on | offIf on instead of doing simple DNS failover (like above), SER will load balance requests to different SRV records of the same priority based on the SRV records weights (like described in RFC 2782). For a destination which has different priorities for all its SRV records, this option will be equivalent with simple DNS failover.
dns_try_ipv6 = on | offShared with the resolver (see resolver description).
dns_try_naptr = on | offShared with the resolver (see resolver description).
dns_udp_pref = number
Shared with the resolver (see resolver description).
dns_tcp_pref = number
Shared with the resolver (see resolver description).
dns_tls_pref = numberShared with the resolver (see resolver description).
dns_cache_flags = dns cache specific resolver flagsUsed for overriding the default behavior (low level).
|1||IPv4 only||Only DNS A requests are performed, even if SER listens also on IPv6 addresses.|
|2||ipv6 only||Only DNS AAAA requests are performed. Ignored if dns_try_ipv6 is off or SER doesn’t listen on any ipv6 address.|
|4||prefer ipv6||Try first to resolve a host name to an IPv6 address (DNS AAAA request) and only if this fails try an IPv4 address (DNS A request). By default the IPv4 addresses are preferred.|
dns_cache_negative_ttl = timeTime to live for negative results (”not found”) in seconds. Use 0 to disable.
dns_cache_min_ttl = timeMinimum accepted time to live for a record, in seconds. If a record has a lower TTL, its value will be discarded and dns_cache_min_ttl will be used instead.
dns_cache_max_ttl = timeMaximum accepted time to live for a record, in seconds. If a record has a higher TTL, its value will be discarded and dns_cache_max_ttl will be used instead.
dns_cache_mem = timeMaximum memory used for the dns cache in Kb.
dns_cache_gc_interval = timeHow often (in s) the DNS cache will be garbage collected.
If defined the DNS cache support will be compiled in (default). If not needed/wanted the dns_cache can be disabled from the ser’s config file. The only advantages for not compiling the DNS cache support is a slight decrease of the executable size and an extremely small performance increase (1 less comparison per dns request).
If defined the DNS failover support will be compiled in. (default). Compiling the DNS failover support has a few disadvantages, see the “Drawbacks” section.
If defined (default) support for load balancing using SRV records weights (as described in RFC 2782) will be compiled in. Note however that it still must be enabled from the SER config, it’s disabled by default (see the dns_srv_lb config option).
(shared with the resolver) if defined NAPTR support will be compiled in (default). Note that even if compiled, NAPTR support must be enabled also from the SER config (see the dns_try_naptr option).
If defined all the additional records in a NAPTR answer will be cached. Normally SER would cache only “related” records (records that are directly referred), but for answers with lots of A/AAAA records it might happen that not all of the SRV records will fit in the AR section. In this case, without this compile option SER will not cache the un-referred A/AAAA records. By default this option is disabled.
If defined (default), records in the AR section of an answer will be cached only if they are “related” to the query. For example if the query is for a SRV record, A & AAAA records in the AR section will be cached only if there are SRV records pointing to them. This avoids adding possible garbage to the cache. If this option is not defined (experimental), everything in the AR section will be added to the cache.
If defined the DNS cache will be very verbose (it will log lots of messages at the L_DBG level).
<note> To remove a compile options, edit SER’s Makefile.defs and remove it from DEFS list. To add a compile options add it to the make command line, e.g.:
make proper; make all extra_defs=-DUSE_DNS_FAILOVER
or for a permanent solution, edit
Makefile.defs and add it to DEFS (don’t forget to prefix it with -D). Some options require editing dns_cache.c or resolve.[ch] (just grep after them).