class BGPMain Next hop resolvability and IGP distances are accessed through this class. More... Definition #include <main.hh> List of all Methods

Annotated List Files Globals Hierarchy Index

Public Methods

•  BGPMain ()
•  ~BGPMain ()
• ProcessStatus  status (string& reason)
• void  local_config (const uint32_t& as, const IPv4& id)
• void  attach_peer (BGPPeer *p)
• void  detach_peer (BGPPeer *p)
• BGPPeer * find_peer (const Iptuple& search)
• bool  create_peer (BGPPeerData *pd)
• bool  delete_peer (const Iptuple& iptuple)
• bool  enable_peer (const Iptuple& iptuple)
• bool  disable_peer (const Iptuple& iptuple)
• bool  next_hop_rewrite_filter (const Iptuple& iptuple, const IPv4& next_hop)
• bool  get_peer_list_start (uint32_t& token)
• bool  get_peer_list_next (const uint32_t& token, IPv4& local_ip, uint32_t& local_port, IPv4& peer_ip, uint32_t& peer_port)
• bool  get_peer_id (const Iptuple& iptuple, IPv4& peer_id)
• bool  get_peer_status (const Iptuple& iptuple, uint32_t& peer_state, uint32_t& admin_status)
• bool  get_peer_negotiated_version (const Iptuple& iptuple, int32_t& neg_version)
• bool  get_peer_as (const Iptuple& iptuple, uint32_t& peer_as)
• bool  get_peer_msg_stats (const Iptuple& iptuple, uint32_t& in_updates, uint32_t& out_updates, uint32_t& in_msgs, uint32_t& out_msgs, uint16_t& last_error, uint32_t& in_update_elapsed)
• bool  get_peer_established_stats (const Iptuple& iptuple, uint32_t& transitions, uint32_t& established_time)
• bool  get_peer_timer_config (const Iptuple& iptuple, uint32_t& retry_interval, uint32_t& hold_time, uint32_t& keep_alive, uint32_t& hold_time_configured, uint32_t& keep_alive_configured, uint32_t& min_as_origination_interval, uint32_t& min_route_adv_interval)
• bool  register_ribname (const string& name)
• void  main_loop ()
• void  terminate ()
• bool  run ()
• int  create_listener (const Iptuple& iptuple)
• LocalData * get_local_data ()
• void  start_server (const Iptuple& iptuple)
• void  stop_server (const Iptuple& iptuple)
• void  stop_all_servers ()
• bool  add_route (const OriginType origin, const AsNum& as, const IPv4& next_hop, const IPv4Net& nlri)
• bool  delete_route (const IPv4Net& nlri)
• bool  get_route_list_start4 (uint32_t& token)
• bool  get_route_list_start6 (uint32_t& token)
• bool  get_route_list_next4 ( const uint32_t& token, IPv4& peer_id, IPv4Net& net, uint32_t& origin, vector<uint8_t>& aspath, IPv4& nexthop, int32_t& med, int32_t& localpref, int32_t& atomic_agg, vector<uint8_t>& aggregator, int32_t& calc_localpref, vector<uint8_t>& attr_unknown, bool& best)
• bool  get_route_list_next6 ( const uint32_t& token, IPv4& peer_id, IPv6Net& net, uint32_t& origin, vector<uint8_t>& aspath, IPv6& nexthop, int32_t& med, int32_t& localpref, int32_t& atomic_agg, vector<uint8_t>& aggregator, int32_t& calc_localpref, vector<uint8_t>& attr_unknown, bool& best)
• bool  rib_client_route_info_changed4 ( const IPv4& addr, const uint32_t& prefix_len, const IPv4& nexthop, const uint32_t& metric)
• bool  rib_client_route_info_changed6 ( const IPv6& addr, const uint32_t& prefix_len, const IPv6& nexthop, const uint32_t& metric)
• bool  rib_client_route_info_invalid4 ( const IPv4& addr, const uint32_t& prefix_len)
• bool  rib_client_route_info_invalid6 ( const IPv6& addr, const uint32_t& prefix_len)
• bool  set_parameter ( const Iptuple& iptuple, const string& parameter)
• BGPPlumbing * plumbing () const
• XrlStdRouter * get_router ()
• EventLoop&  eventloop ()
• XrlBgpTarget * get_xrl_target ()
• void  notify_birth (const string& target_class, const string& target_instance)
• void  notify_death (const string& target_class, const string& target_instance)
• bool  processes_ready ()
• string  bgp_mib_name () const
• bool  do_snmp_trap () const
• void  finder_death ()

Detailed Description

Next hop resolvability and IGP distances are accessed through this class.

Next hop resolvability and IGP distances are retrieved from the RIB and cached here in BGP. This retrieval process implicitly registers interest with the RIB regarding these next hops. Thus any changes in these next hops is signalled by the RIB to BGP via callbacks.

If the state of a next hop changes (resolvable/unresolvable), or an IGP distance changes, then it is possible that a new route may now win the decision process. The decision process must therefore be re-run for all routes that are affected by a next hop change. This re-run of the decision process is achieved calling "igp_nexthop_changed" on the decision process.

What questions can be asked about next hops? Is a next hop resolvable and if it is, what is the IGP distance.

To answer questions about next hops three interfaces are supported: 1) An asynchronous interface that registers a callback which will be called when a response becomes available. For use by the (next hop) route table before decision. By the time a route gets to decision it *must* be known if the route is resolvable. 2) A synchronous interface for use by decision. It is a fatal error if this interface is called and the next hop is not in the cache. As by the time decision is called the cache should have been populated by use of the asynchronous interface. 3) A synchronous debugging interface.

Cache maintainance: Every stored SubnetRoute in every rib in has a next hop. Every unique next hop has an entry in the cache. If a next hop lookup through the asynchronous interface causes a cache miss then an entry is created with a reference count of 1. Subsequent lookups through the next hop interface will cause the reference count to be incremented by 1. An interface to increase the reference count by more than one also exists. All route deletions should explicitly call a routine in here to decrement the reference count.

Get the process status

Set the local configuration.

Parameters:

as	as number.
id	router id.

attach peer to peerlist

Parameters:

p

BGP peer.

detach peer from the peerlist.

Parameters:

p

BGP peer.

Find peer with this iptuple

Parameters:

search

iptuple.

Returns: A pointer to a peer if one is found NULL otherwise.

create a new peer and attach it to the peerlist.

Parameters:

pd	BGP peer data.

Returns: true on success

delete peer tear down connection and remove for peerlist.

XrlBgpTarget xbt(bgp.get_router(), bgp);

Parameters:

iptuple

iptuple.

Returns: true on success

enable peer

Parameters:

iptuple

iptuple.

Returns: true on success

disable peer

Parameters:

iptuple

iptuple.

Returns: true on success

shutdown BGP cleanly

Stop listening for incoming connections.

add route

Typically called via XRL's to add static routes.

Parameters:

origin	the origin of the path information.
asum	an as number will eventually be a path.
next_hop	the border router that should be used as a detination for the nlri.
nlri	network level reachability information.
success	returns true if the route was sucessfully added.

delete route

Typically called via XRL's to delete static routes.

Parameters:

nlri	network level reachability information.
success	returns true if the route was sucessfully added.

set parameter

Typically called via XRL's to set which parameters we support per peer.

Parameters:

iptuple	iptuple
parameter	we are setting for this peer.

[const]

Originally inserted for testing. However, now used by all the "rib_client_route_info_*" methods.

Call via XrlBgpTarget when the finder reports that a process has started.

Parameters:

target_class	Class of process that has started.
target_instance	Instance name of process that has started.

Call via XrlBgpTarget when the finder reports that a process has terminated.

Parameters:

target_class	Class of process that has terminated.
target_instance	Instance name of process that has terminated.

Returns: Return true when all the processes that BGP is dependent on are ready.

[const]

Returns: Return the bgp mib name.

[const]

Check to see if the bgp snmp entity is running.

To be called when the finder dies.