class RoutingPlatf;
/** @ingroup SURF_routing_interface
- * @brief A network card
+ * @brief Network cards are the vertices in the graph representing the network, used to compute paths between nodes.
+ *
* @details This represents a position in the network. One can route information between two netcards
*/
class NetCard {
public:
virtual ~NetCard(){};
- virtual int getId()=0;
+ virtual int getId()=0; // Our rank in the vertices_ array of our englobing AS.
virtual int *getIdPtr()=0;
virtual void setId(int id)=0;
virtual char *getName()=0;
As *father_ = nullptr;
xbt_dict_t sons_ = xbt_dict_new_homogeneous(NULL);
- xbt_dynar_t p_indexNetworkElm = xbt_dynar_new(sizeof(char*),NULL); // TODO: What is it?
+ xbt_dynar_t vertices_ = xbt_dynar_new(sizeof(char*),NULL); // our content, as known to our graph routing algorithm (maps vertexId -> vertex)
xbt_dict_t bypassRoutes_ = nullptr;
e_surf_routing_hierarchy_t hierarchy_ = SURF_ROUTING_NULL;
xbt_dynar_t upDownLinks = xbt_dynar_new(sizeof(s_surf_parsing_link_up_down_t),NULL);
/**
- * @brief Get the characteristics of the routing path between two points
+ * @brief Probe the routing path between two points
*
- * This function is used by the networking model to find the information it needs when starting a communication.
+ * The networking model uses this function when creating a communication
+ * to retrieve both the list of links that the create communication will use,
+ * and the summed latency that these links represent.
*
- * The things are not straightforward because the platform can be routed using several routing models.
+ * The network could recompute the latency by itself from the list, but it would
+ * require an additional link set traversal. This operation being on the critical
+ * path of SimGrid, the routing computes the latency in behalf of the network.
+ *
+ * Things are rather complex here because we have to find the path from ASes to ASes, and within each.
+ * In addition, the different ASes may use differing routing models.
* Some ASes may be routed in full, others may have only some connection information and use a shortest path on top of that, and so on.
* Some ASes may even not have any predefined links and use only coordinate informations to compute the latency.
*
*
* @param src Initial point of the routing path
* @param dst Final point of the routing path
- * @param into Container into which the links should be pushed
- * @param latency Accumulator in which the latencies should be added
+ * @param into Container into which the traversed links should be pushed
+ * @param latency Accumulator in which the latencies should be added (caller must set it to 0)
*/
- virtual void getRouteAndLatency(
- NetCard *src, NetCard *dst,
- sg_platf_route_cbarg_t into, double *latency)=0;
+ virtual void getRouteAndLatency(NetCard *src, NetCard *dst, sg_platf_route_cbarg_t into, double *latency)=0;
+ /** @brief retrieves the list of all routes of size 1 (of type src x dst x Link) */
virtual xbt_dynar_t getOneLinkRoutes()=0;
virtual void getGraph(xbt_graph_t graph, xbt_dict_t nodes, xbt_dict_t edges)=0;
