Provided by: gdnsd_1.11.1-1_amd64 bug

NAME

       gdnsd-plugin-multifo - gdnsd plugin for multi-address, all-active failover groups

SYNOPSIS

       Example plugin config:

         plugins => {
           multifo => {
             up_thresh => 0.3
             service_types => default,
             v4www => {
               lb01 => 192.0.2.200,
               lb02 => 192.0.2.201,
               lb03 => 192.0.2.202,
             }
             v6smtp => {
               service_types => [ smtp ],
               up_thresh => 0.1,
               lb01_v6 => 2001:DB8::1,
               lb02_v6 => 2001:DB8::2,
               lb03_v6 => 2001:DB8::3,
             }
             pubwww => {
               up_thresh => 0.5
               service_types => [ corpwww_type, default ],
               addrs_v4 => [ 192.0.2.100, 192.0.2.101, 192.0.2.102 ]
               addrs_v6 => {
                 service_types => [ up ],
                 up_thresh => 0.7
                 lb01_v6 => 2001:DB8::1,
                 lb02_v6 => 2001:DB8::2,
                 lb03_v6 => 2001:DB8::3,
               }
             }
           }
         }

       Example zonefile RRs:

         web4 180 DYNA multifo!v4www
         smtp 180 DYNA multifo!v6smtp
         www 180 DYNA multifo!pubwww

DESCRIPTION

       gdnsd-plugin-multifo is designed to do multi-address all-active failover grouping.
       Basically, for each configured resource name, you supply a labeled list of addresses.
       multifo monitors these addresses according to "service_types", and answers "DYNA" address
       queries using the non-"DOWN" subset.  The core gdnsd code will round-robin rotate the
       records on the way out, as it does for all address RR-sets.

TOP-LEVEL PLUGIN CONFIG

       At the top level of the plugin's configuration stanza, two special parameters "up_thresh"
       and "service_types" are supported. These set default per-resource options of the same name
       for any resources which do not define them explicitly.

       The rest of the hash entries at the top level are the names of the resources you define.
       Each resource gets a configuration hash of its own for containing resource-specific
       parameters as well as the actual address data.

RESOURCE CONFIG

       Within a resource, you have two basic options.  You can either directly specify a set of
       "label => address" pairs which are all the same family (IPv4 or IPv6), or you can use the
       sub-stanzas "addrs_v4" and/or "addrs_v6" to specify one or both families in the same
       resource.

       The "up_thresh" and "service_types" parameters are inherited through every level, and can
       be overridden at any level (even per-address-family):

       up_thresh
           Floating point, default 0.5, range (0.0 - 1.0].  This configures the per-resource
           "up_thresh" threshold.  More details in "UP THRESH" below.

       service_types
           Array of strings, or single string.  Default "default".  This sets the monitored
           service_types for this resource.  If an array of more than one is provided, all will
           be monitored for each address, and the net monitored state will be the minimum (worst)
           of the set.  See gdnsd.config(8) for more details on service_types.

SHORTCUT CONFIG

       If you have no parameters (service_types, up_thresh) to configure in a given stanza
       (single-family direct resource config, or addrs_v[46]), and do not care about the
       descriptive per-address labels used in monitoring, you can replace the hash with an array
       of addresses.  The labels will be generated for you as a series of integers starting with
       1.  For example, the following are equivalent:

          res1 => { addrs_v4 => [ 192.0.2.1, 192.0.2.2 ] }
          res1 => { addrs_v4 => { 1 => 192.0.2.1, 2 => 192.0.2.2 } }

OPERATIONAL MECHANICS

       All of the addresses for all of the resources are monitored using the per-address-family
       inherited "service_types" specified (default would be simple HTTP/1.0 monitoring for "200
       OK" on port 80).  When the core daemon requests a lookup for address records of a given
       family on one of this plugin's resources, it goes through essentially the following
       process to determine the set of response addresses for that address family: 1) Add all
       non-DOWN addresses to the result set.  2) If the set of non-DOWN addresses fail the
       up_thresh check, add *all* addresses to the result set as a fallback.  3) If any address
       is in the DOWN or even DANGER state, cut the zonefile-specified TTL in half

       This process is repeated independently for each of the IPv4 and IPv6 address subsets, in
       the case that a resource has both address families configured (the TTL is only cut in half
       once of course).  Details on the up_thresh check follow:

UP THRESH

       Any non-DOWN address is considered an acceptable result for up_thresh purposes, even ones
       in the transitional DANGER state.  If there are not "enough" non-DOWN addresses to pass
       the threshold (per address family), all addresses (of a given address family) will be
       returned as a fallback.

       The threshold is implemented mathematically as in the following pseudo-code "if(non_down
       >= ceil(thresh * total)) threshold_passed;".  For example, if thresh is at the default
       value of 0.5, and there are 3 total IPv4 addresses, then 2 of them must be non-down to
       pass the threshold.  The net result is that with the default threshold, the plugin will
       never return an isolated single address from a set of 3.  It will either return all 3, or
       it will return 2/3 if a single address from the set has failed.

       When the threshold check fails (and all addresses are returned) for either address family,
       resource-level total failure will also be signaled to any applicable upstream meta-plugins
       such as metafo or geoip.

       General rules for the results of the up_thresh formula:

       •   A threshold of 1.0 will only pass if all addresses are not-down.  This is mostly
           pointless, you might as well not monitor anything and set up these addresses as a
           static set in a zonefile.

       •   A threshold of 0.01 will pass even if only one address is alive and return just that
           one address, even if it's e.g. the only one left out of 40.

       •   Because a threshold of 0.0 is illegal, if all addresses are down the threshold will
           always fail, returning all addresses.

       Intermediate value examples: (threshold: non-down/total required to pass threshold):

       •   0.1: 1/1 1/2 1/3 1/4 1/5 1/6 1/7 1/8 2/16

       •   0.2: 1/1 1/2 1/3 1/4 1/5 2/6 2/7 2/8 4/16

       •   0.3: 1/1 1/2 1/3 2/4 2/5 2/6 3/7 3/8 5/16

       •   0.4: 1/1 1/2 2/3 2/4 2/5 3/6 3/7 4/8 7/16

       •   0.5: 1/1 1/2 2/3 2/4 3/5 3/6 4/7 4/8 8/16

       •   0.6: 1/1 2/2 2/3 3/4 3/5 4/6 5/7 5/8 10/16

       •   0.7: 1/1 2/2 3/3 3/4 4/5 5/6 5/7 6/8 12/16

       •   0.8: 1/1 2/2 3/3 4/4 4/5 5/6 6/7 7/8 13/16

       •   0.9: 1/1 2/2 3/3 4/4 5/5 6/6 7/7 8/8 15/16

SEE ALSO

       gdnsd.config(5), gdnsd.zonefile(5), gdnsd(8), gdnsd-plugin-simplefo(8)

       The gdnsd manual.

COPYRIGHT AND LICENSE

       Copyright (c) 2012 Brandon L Black <blblack@gmail.com>

       This file is part of gdnsd.

       gdnsd is free software: you can redistribute it and/or modify it under the terms of the
       GNU General Public License as published by the Free Software Foundation, either version 3
       of the License, or (at your option) any later version.

       gdnsd is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without
       even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
       GNU General Public License for more details.

       You should have received a copy of the GNU General Public License along with gdnsd.  If
       not, see <http://www.gnu.org/licenses/>.