| Age | Commit message (Collapse) | Author | |
|---|---|---|---|
| 2005-12-29 | More correct cleanup on exit. OK norby@ | Claudio Jeker | |
| 2005-12-29 | Instead of a simple timer per neighbor for the LS retransmition list use | Claudio Jeker | |
| a timewheel. All LSA updates are now first added to the queue and sent out later. This makes it possible to cluster multiple LSA updates into on OSPF packet. This gives a massive speedup when large databases need to be synced. Tested and OK norby@ | |||
| 2005-12-15 | No need to call the nbr_fsm for unexpected LS requests. The RFC tells to | Claudio Jeker | |
| just ignore thos packets. OK norby@ | |||
| 2005-12-15 | Fix ospfe shutdown function. The cleanup was totaly busted. OK norby@ | Claudio Jeker | |
| 2005-12-15 | Simpify iface_del() and nbr_del(). make them void funtions as they can not fail | Claudio Jeker | |
| and do not remove a element twice in iface_del(). OK norby@ | |||
| 2005-12-15 | Fix memory leaks in the parser. Make the string passed with area more strict. | Claudio Jeker | |
| Finaly implement clear_config(). OK norby@ | |||
| 2005-12-14 | Replace strlcpy() with memcpy() in the sockaddr_dl handling. sdl_data is | Claudio Jeker | |
| not a C string and strlcpy() only works on C strings -- returns length of the source. Found by tedu@ OK deraadt@ tedu@ | |||
| 2005-12-05 | Optimize rde_redistribute() a bit and remove two unneeded and way to verbose | Claudio Jeker | |
| log_debug() messages. OK norby@ | |||
| 2005-12-04 | Increase ADJ_TMOUT to prevent loss of adjacency during bulk ls update | Esben Norby | |
| transfers. ok claudio@ | |||
| 2005-12-04 | Increase the receive buffer of the raw socket. | Esben Norby | |
| Prevents some of the retransmitting during bulk ls update transfers. ok claudio@ | |||
| 2005-12-02 | Fixed version of r1.35, readd candidate to list in both cases so that the list | Claudio Jeker | |
| remains sorted. Tested and OK norby@ | |||
| 2005-11-14 | Revert candidate list change since it is broken. | Esben Norby | |
| ok henning@ | |||
| 2005-11-12 | spacing mostly | Theo de Raadt | |
| 2005-11-04 | The candidate list is a sorted linear list so when changing the cost of | Claudio Jeker | |
| a vertex that is already a candidate it is necessary to remove and insert the candidate otherwise the candidate list is no longer correctly sorted. OK norby@ | |||
| 2005-11-04 | The maximal rate LSA can be updated is all MIN_LS_INTERVAL seconds and not | Claudio Jeker | |
| MIN_LS_ARRIVAL. MIN_LS_ARRIVAL is used to limit the rate of incomming updates. | |||
| 2005-11-04 | Set the vertex timestamp as soon as possible giving ospfd a chance to recover | Claudio Jeker | |
| from time jumps. | |||
| 2005-11-04 | Kill spf_timer in struct ospfd_conf. There is no need for two event structs | Claudio Jeker | |
| just use ev for the spf_timer -- ev is unused in the RDE. | |||
| 2005-11-04 | Use >= instead of == to compare cost with LS_INFINITY. While there fix a typo. | Claudio Jeker | |
| 2005-11-04 | In the super fast start-up case in recv_db_description() make sure the | Claudio Jeker | |
| the 2WAY neighbor event is issued before the interface NBR_CHNG. Fix some comments while there. | |||
| 2005-11-04 | Remove unused variable link_state in struct nbr. | Claudio Jeker | |
| 2005-11-04 | Make "interface em0 { }" work again. Reported and OK stevesk@ | Claudio Jeker | |
| 2005-10-30 | remove some unneeded #includes and correct keyword in error string; | Kevin Steves | |
| ok claudio@ | |||
| 2005-10-28 | a little bit saner macros; | Jason McIntyre | |
| 2005-10-28 | authentication cleanup and clarification; | Kevin Steves | |
| ok and mdoc help claudio@ jmc@ | |||
| 2005-10-28 | clean auth error messages; ok claudio@ msf@ | Kevin Steves | |
| 2005-10-27 | Add something to explain effect of passive interface; ok claudio@ | Kevin Steves | |
| 2005-10-26 | some unneeded #includes; ok claudio@ | Kevin Steves | |
| 2005-10-25 | 0) ospfctl doesn't use ospfd.conf so remove that from FILES | Kevin Steves | |
| 1) add /var/run/ospfd.sock to ospfd and ospfctl FILES 2) add .Xr ospfd.conf 5 to ospfctl SEE ALSO ok claudio@ jmc@ | |||
| 2005-10-21 | htons() instead of ntohs() when going from host to network. | Claudio Jeker | |
| 2005-10-21 | Revert part of rev 1.14. For now queue LSAs to all neighbors if the interface | Claudio Jeker | |
| is in state BACKUP or DR -- even to the DR. This causes unneeded retransmission of LSAs but solves a problem with self originated ones. The BDR was no longer capable to reliably flood self originated LSA. Better solution will follow. Reported by stevesk@ | |||
| 2005-10-19 | small cleanups while reading; ok claudio@ | Kevin Steves | |
| 2005-10-19 | Add neighbor ID in warning and use correct function names. | Claudio Jeker | |
| 2005-10-19 | fix some spellings in comments. | Claudio Jeker | |
| 2005-10-19 | header cleaning; ok claudio@ | Kevin Steves | |
| 2005-10-19 | use string.h; ok claudio@ | Kevin Steves | |
| 2005-10-18 | Simplify code a bit. OK norby@ | Claudio Jeker | |
| 2005-10-18 | Cleanup interface code a minimal bit. There is no need for if_shutdown(), | Claudio Jeker | |
| especially it got called in the wrong place. OK norby@ | |||
| 2005-10-18 | Better warning for non equal MTU in the database phase. Instead of just | Claudio Jeker | |
| whining that the MTU differs tell actually what is expected and what was sent. OK norby@ | |||
| 2005-10-18 | The kroute nexthop stuff was removed long time ago. So remove these | Claudio Jeker | |
| function prototypes too. | |||
| 2005-10-18 | Correct log_debug(). Found while scrolling through that file. | Claudio Jeker | |
| 2005-10-13 | Fix a memory leak in the ospf engine. First of all ospfd did not track if a | Claudio Jeker | |
| LSA was already queued on nbr self. So after the first round -- where the lsa was queued -- ospfd would remove it again in a second run. That was not enough but the LSA was removed from the wrong neighbor -- nbr instead of iface->self. LIST_REMOVE does not track the queue and so that resulted in some strangeness (especially the lsa_cache did not get cleared). OK norby@ | |||
| 2005-10-13 | Correctly queue ls updates on the retransmission lists. | Claudio Jeker | |
| First of all it is not clever to set queued before we actually queue something -- unless queued is set. If ospfd is running as BDR don't add updates to the retransmission list to the DR. OK norby@ | |||
| 2005-10-12 | Correct function name in warning. | Claudio Jeker | |
| 2005-10-12 | Revert last commit. That file was not intended to be commited. | Claudio Jeker | |
| 2005-10-12 | Fix the flooding procedure. Violate the RFC else many BAD_LS_REQ events and | Claudio Jeker | |
| session drops happend while booting up. If a router is conected to a other router over two different interface one session will be unable to load until the other one is in state FULL. ospfd no longer issues a BAD_LS_REQ event if the LSA is equal to the one in table but if the sent lsa is older a BAD_LS_REQ event is still issued. OK norby@ | |||
| 2005-10-12 | Fix a massive memory leak in the RDE. ospfd leaked every ls update unless | Claudio Jeker | |
| lsa_add() was called (which is just one specific case of many) OK norby@ | |||
| 2005-10-12 | If router is DR or BDR join the AllDRouter list (and leave it if otherwise). | Claudio Jeker | |
| OK norby@ | |||
| 2005-10-12 | Minor cleanup. Use iface->state & IF_STA_DRORBDR as used in other places | Claudio Jeker | |
| instead of iface->state != (IF_STA_DR | IF_STA_BACKUP). OK norby@ | |||
| 2005-10-12 | In one special condition ospfd sent out ls updates without updating the | Claudio Jeker | |
| checksum. Because of the invalid checksum the other routers dropped it and so the update was retransmitted over and over again. Now we update the checksum in lsa_merge() after bumping the sequence number instead of hoping that lsa_timeout() -> lsa_refresh() will do it for us. OK norby@ | |||
| 2005-10-06 | I should have read the RFC more carefully. | Claudio Jeker | |
| ...if only one of the instances has its LS age field set to MaxAge The important part is "only one", so check if both LSA are at MaxAge and in that case return 0 -- the LSAs are identical. This fixes a super nasty doom loop between two ospfd sending each other LSA updates at an incredible speed. Tested and OK norby@ | |||
 |
index : src | |
| OpenBSD base system |
| summaryrefslogtreecommitdiff |