summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2010-12-08ospfd: Fix maxage/flush to not try flood twice, remember maxages for longerPaul Jakma
2006-05-30 Paul Jakma <paul.jakma@sun.com> * (general) Fix confusion around MaxAge-ing and problem with high-latency networks. Analysis and suggested fixes by Phillip Spagnolo, in [quagga-dev 4132], on which this commit expands slightly. * ospf_flood.{c,h}: (ospf_lsa_flush) new function. Scope-general form of existing flush functions, essentially the dormant ospf_maxage_flood() but without the ambiguity of whether it is responsible for flooding. * ospf_lsa.c: (ospf_lsa_maxage) Role minimised to simply setup LSA on the Maxage list and schedule removal - no more. ospf_lsa_flush* being the primary way to kick-off flushes of LSAs. Don't hardcode the remover-timer value, which was too short for very high-latency networks. (ospf_maxage_lsa_remover) Just do what needs to be done to remove maxage LSAs from the maxage list, remove the call to ospf_flood_through(). Don't hardcode remove-timer value. (ospf_lsa_{install,flush_schedule}) ospf_lsa_flush is the correct entrypoint to flushing maxaged LSAs. (lsa_header_set) Use a define for the initial age, useful for testing. * ospf_opaque.c: (ospf_opaque_lsa_refresh) ditto. (ospf_opaque_lsa_flush_schedule) ditto. * ospfd.h: ({struct ospf,ospf_new}) Add maxage_delay parameter, interval to wait before running the maxage_remover. Supply a suitable default. Add a define for OSPF_LSA_INITIAL_AGE, see lsa_header_set().
2010-12-08ospfd: Unify router and network LSA refresh logic with general refresherPaul Jakma
* (general) Get rid of the router and network LSA specific refresh timers and make the general refresher do this instead. Get rid of the twiddling of timers for router/network LSA that was spread across the code. This lays the foundations for future, general LSA refresh improvements, such as making sequence rollover work, and having generic LSA delays. * ospfd.h: (struct ospf) Bye bye to the router-lsa update timer thread pointer. (struct ospf_area) and to the router-lsa refresh timer. * ospf_interface.h: Remove the network_lsa_self timer thread pointer * ospf_lsa.h: (struct ospf_lsa) oi field should always be there, for benefit of type-2/network LSA processing. (ospf_{router,network}_lsa_{update_timer,timer_add}) no timers for these more (ospf_{router,network}_lsa_update) more generic functions to indicate that some router/network LSAs need updating (ospf_router_lsa_update_area) update router lsa in a particular area alone. (ospf_{summary,summary_asbr,network}_lsa_refresh) replaced by the general ospf_lsa_refresh function. (ospf_lsa_refresh) general LSA refresh function
2010-12-08ospfd: Remember network LSA sequence numbers across up/downs of an interfacePaul Jakma
* ospf_interface.h: (struct ospf_if_params) add field for saved network LSA seqnum * ospf_interfa.c: (ospf_new_if_params) init network_lsa_seqnum field to initial seqnum - doesnt matter though. * ospf_lsa.c: (ospf_network_lsa_new) check for any saved sequence number, and use if it exists. Save the result back. This should help avoid needless round of LSUpdate/LSRequests when a neighbour has to tell the originator "uhm, i have something newer than that already". * ospf_vty.c: (show_ip_ospf_interface_sub) Show the saved network LSA seqnum
2010-12-08ospfd: Prioritise hellos for sending by queueing to head of output bufferPaul Jakma
* It's possible for the packet output buffer to be filled up with a long series of non-Hello packets in between Hellos packets, such that the router's neighbours don't receive the Hello packet in time, even though the hello-timer ran at about the right time. Fix this by prioritising Hello packets, letting them skip the queue and go ahead of any packets already on the queue. This problem can occur when there are lots of LSAs and slow links. * ospf_packet.h: (ospf_hello_send_sub) not used outside of ospf_packet.c * ospf_packet.c: (ospf_fifo_push_head) add packet to head of fifo (so its no longer really a fifo, but hey) (ospf_packet_add_top) add packet to top of the packet output queue. (ospf_hello_send_sub) Put Hello's at the top of the packet output queue. make it take in_addr_t parameter, so that this ospf_hello_send can re-use this code too. (ospf_hello_send) consolidate code by using ospf_hello_send_sub (ospf_poll_send,ospf_hello_reply_timer) adjust for ospf_hello_send_sub.
2010-12-08ospfd: Reset neighbour inactivity timer for any packet arrivalPaul Jakma
* The hello protocol monitors connectivity in 2 different ways: a) local -> remote b) remote -> local Connectivity is required in both directions (2-way) for adjacencies to form. The first requires a round-trip to detect, and is done by advertising which other hosts a router knows about in its hello messages. This allows a host to detect which other routers are and are not receiving its message. If a remote neighbour delists the local router, then the local router raises a "1-Way Received" event. The latter is straight-forward, and is detected by setting a timer for the neighbour. If another Hello packet is not received within this time then the neighbour is dead, and a separate "Inactive" event is raised. These are 2 different and relatively independent measures. Knowing that we can optimise the 2nd, remote->local measure and reset the timer when /any/ packet arrives from that neighbour. For any packet is as good as a Hello packet. This can help in marginal situations, where the number of protocol messages that must be sent sometimes can exceed the capacity of the network to transmit the messages within the configured dead-time. I.e. an OSPF network with lots of LSAs, slow links and/or slow hosts (e.g. O(10k) LSAs, O(100kbit) links, embedded CPUs, and O(10s) dead-times). This optimisation allows an OSPF network to run closer to this margin, and/or allows networks to perhaps better cope with rare periods of exceptional load, where otherwise they would not. It's fully compatible with plain OSPF implementations and doesn't prejudice dead-neighbour detection. * ospf_nsm.h: Rename HelloReceived event to PacketReceived. * ospf_nsm.c: (nsm_hello_received) -> nsm_packet_received * ospf_packet.c: Schedule PacketReceived whenever a valid message is received.
2010-12-08ospfd: the maxage_lsa_remover should check whether it needs to yield the cpuPaul Jakma
2010-12-08ospfd: Fix various route_unlock discrepenciesPaul Jakma
* ospf_ase.c: (ospf_ase_calculate_route) Fix compiler warning about eval needing brackets. (various) add defensive asserts. * ospf_lsdb.c: (ospf_lsdb_add) add missing node unlock if same lsa already was indexed. (ospf_lsdb_delete) check it's actually the same as specified lsa before deleting (ospf_lsdb_lookup_by_id_next) fix another corner case - no result => don't go on.
2010-12-08ospfd: fix lsa_refresh_walker unlock before use bugPaul Jakma
* ospf_lsa.c: (ospf_lsa_refresh_walker) fix an "unlock before use" bug (various) add asserts for lsa refcounting.
2010-12-08ospfd: interface code should leave network_lsa_self alonePaul Jakma
* ospf_interface.c: (ospf_if_{new,cleanup}) don't touch the network_lsa_self, ISM and NSM take care of cleaning it up if needs be + we want to keep network_lsa_self around when possible for the the seqnum. This shouldn't really make much difference though, particularly as we have a separate sequence number memory mechanism.
2010-12-08ospfd: OSPF_MIN_LS_ARRIVAL compare should be >= to match ospf_floodPaul Jakma
* ospf_packet.c: (ospf_ls_upd) the corresponding test on the arrival side in (ospf_flood) is <, so this should be >=, not >, purely for consistency. There is no practical effect here though.
2010-12-08ospfd: ospf_if_free can leave dangling references on ISM events - cancel themPaul Jakma
* ospf_interface.c: (ospf_if_free) events with dangling pointers left scheduled can be seriously bad for ospfd's health. Cancel the event.
2010-12-08ospfd: Lower level of some common messages from info to debugPaul Jakma
* ospf_{ism,network}.c: Certain oft-repeated but trivial messages should be debug log level, not info, to avoid spamming 'terminal monitor'
2010-12-08lib: Fix bug in prefix trie lookupPaul Jakma
* lib/table.c: (route_node_match) fix overshoot that was causing this function to go 1 bit too far and thus reading past end of prefix. (route_node_lookup) be defensive - don't assume others will clean up leaves when removing info.
2010-12-08lib: prefix.c nano-optimisationPaul Jakma
* lib/prefix.c: (prefix_match) nano-optimisation, let it return early without copying pointers.
2010-12-08lib: Make workqueue more conservative about ramping upPaul Jakma
* workqueue.c: (work_queue_run) Err more on the side of keeping granularity down, by being more conservative about increasing it. Also, fix mispelling.
2010-12-08lib: Add a command to clear the thread CPU history dataPaul Jakma
* (general) this can be useful when investigating thread latency problems, when you don't want to have to restart a daemon between tests. * thread.c: (cpu_record_(hash_)clear) wipe the stored thread cpu history data, according to the filter, similar to the vty print code. (clear_thread_cpu_cmd) new command to clear data. * thread.h: export new command * command.c: install it
2010-12-08lib: Thread scheduler should be fair and not let events starve I/O and timersPaul Jakma
* thread.c: (thread_fetch) the current scheduler will service events indefinitely, ignoring I/O and timers, so long as there are events. In other words, events can crowd out I/O and timers. In theory this shouldn't be a huge problem as events are generated only by timers and I/O, however in practice it means normal-load behaviour is not as useful a predictor of high-load behaviour as it should be. Fix this by considering all the kinds of threads, in every run of the scheduler. For any given run, we prioritise events, however across runs the scheduler should be fair. This has been observed to give more stable inter-packet times in testing of ospfd (i.e. lower std-dev). (thread_process) new heler to queue all the given threads onto the ready list
2010-12-08lib: thread history funcname shouldn't be constant, it's freedPaul Jakma
* thread.h: (struct cpu_thread_history.funcname) malloc gets called on this so it should not be const * thread.c: (cpu_record_print) cast the (const char *) to (char *), this function knows what it's doing (free wont be called on it).
2010-12-08bgpd: Simplify process queue initPaul Jakma
* bgp_route.c: (bgp_process_queue_init) rsclient wasn't getting all the same things initialised as the main queue. Simplify to make it more robust.
2010-11-03infrastructure: Express preference for published git reposGreg Troxel
* HACKING: Express notion that a published git repository is preferred. Fold request for commit message into patch section. Express desire for comments in code explaining correctness of post-commit state, and for commit message to explain correctness of the change.
2010-11-03infrastructure: Whitespace cleanup.Greg Troxel
HACKING: Whitespace changes only.
2010-11-03infrastructure: Clarify commit message format.Greg Troxel
Note 54/72 line length rules, and that this is intended to play nice with "git log --oneline".
2010-11-03Adjust description of tomhenderson repo.Greg Troxel
Tom's repo is now a clone of the official repo and has an mttr branch off of 0.99.17.
2010-10-21bgpd: fix printed value of last-update timestampVladimir L Ivanov
* bgp_route.c: (route_vty_out_detail) calculate time value in a way, which works regardless of monotonic clock being used or not
2010-10-18zclient: fix router-id calculation for IPv6 (#595)Dmitry Tejblum
If router-id is not specified in ospf6d.conf, ospf6d will get it from the zebra daemon. But ospf6d originates Link LSAs before the router-id is returned by zebra, thus this router's Link LSAs will be flooded with AdvRouter set to 0. * zclient.c: zclient_start(): send ZEBRA_INTERFACE_ADD message after ZEBRA_ROUTER_ID_ADD, not before
2010-10-06bgpd, lib: adopt afi_t and safi_t in several placesMichael Lambert
* bgpd/bgp_attr.c, bgpd/bgp_open.h, bgpd/bgp_route.c, lib/prefix.c, lib/prefix.h: Various integer types were being used where, if we had strict type checking, afi_t and safi_t would be required. Signed-off-by: G.Balaji <balajig81@gmail.com> (cherry picked from commit c8af35ffa2dc79ff7d7ff00b1b61f1f50a100ab6)
2010-09-24Add MTR repository.Greg Troxel
2010-09-17lib/vty.c: add missing format string when printing out motd messageNico Golde
Signed-off-by: G.Balaji <balajig81@gmail.com>
2010-09-17Document rules for zalloc and friends.Greg Troxel
lib/memory.c:z{a,c,re}alloc, zfree, zdup: add requires/effects comments.
2010-09-17Add pointer to Balaji G. git.Greg Troxel
Balaji has been accumulating patches from the mailing list and pre-screening them and spiffing up commit messages.
2010-09-17Set from even if binfo->extra is NULL.Greg Troxel
bgpd/bgp_packet.c:bgp_update_packet(): When extracting the peer, don't fail to extract it because "binfo->extra" is NULL. While one should certainly avoid dereferencing binfo->extra, that's not a good reason not to use binfo->peer. Fixes https://bugzilla.quagga.net/show_bug.cgi?id=497. Patch by Eric Sobocinksi.
2010-08-19Zebra zserv: bogus conditionalStephen Hemminger
This looks like a bug in original code from misunderstanding of C rules of evaluation.
2010-08-19release: 0.99.17Denis Ovsienko
* configure.ac: bump up version number
2010-08-09bgpd: fix handling of AS path dataChris Hall
* bgpd/bgp_aspath.c * assegments_parse(): add handling of AS4_PATH input, update bounds checks, add check for AS segment type * aspath_parse(): add handling of AS4_PATH input, expect assegments_parse() to do length checking * aspath_empty(): update for the new function prototype * bgpd/bgp_aspath.h: ditto * tests/aspath_test.c: ditto * bgpd/bgp_attr.c * bgp_attr_aspath(): add handling of AS4_PATH input, update flags checks, change returned type * bgp_attr_as4_path(): discard, superseded by bgp_attr_aspath() * bgp_attr_parse(): update respectively
2010-05-14bgpd: tighten bounds checking in RR ORF msg readerChris Hall
* bgp_packet.c: (bgp_route_refresh_receive) add validation of "Length" (RFC5292) field value, check input stream bounds each time bytes are pulled from it
2010-05-05zebra: fix infinite loop when deleting an interfaceRoman Hoog Antink
When deleting a VLAN interface after flushing its addresses, zebra uses 100% CPU time and freezes. * interface.c: The while loop in line 407 that should clean up connected routes never hits one of the 2 lines "last = node;" and thus loops forever. Signed-off-by: Roman Hoog Antink <rha@open.ch>
2010-05-05Update for git and emphasize asking for good reports.Greg Troxel
2010-04-28build: ignore mkinstalldirs and texinfo.texDavid Lamparter
these are autogenerated by autoreconf, drop them from git.
2010-04-28isisd: change ISIS_METHOD to use C preprocessorDavid Lamparter
this fixes warnings from vtysh extract.pl by making sure the isis method files always 'work'. (previously, extract.pl would grab unselected isis method sources and then complain about missing headers)
2010-04-28ospf6d: Fix crash when '[no] ipv6 ospf6 advertise prefix-list' is in ↵David Ward
startup-config * ospf6_interface.c: When '[no] ipv6 ospf6 advertise prefix-list' appears in the startup configuration for ospf6d, a crash occurs, because ospf6d attempts to schedule LSAs when the 'oi->area' structure has not yet been initialized. Now, when the command above is issued (either in the startup configuration or at runtime), ospf6d will only schedule LSAs if the 'oi->area' structure has been initalized. A similar test is already used when handling the commands 'ipv6 ospf6 priority' and 'ipv6 ospf6 cost'.
2010-04-20build/extra: Enhance README.NetBSD make/gmake decision.Greg Troxel
2010-04-18ospfd: Only refresh external default route once.Joakim Tjernlund
* ospf_zebra.c: (ospf_distribute_list_update_timer) forces a refresh of default route each time it finds a default prefix. This is suboptimal, just record that it needs to be done and do it once. Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
2010-04-18ospfd: Make sure ospf_distribute_list_update_timer() eventually runs.Joakim Tjernlund
* ospf_zebra.c: (ospf_distribute_list_update_timer) If there are updates to the distribute list every 5 second or less, ospf_distribute_list_update_timer() will never run as the timer gets rearmed for each update. This fixes it by never rearming an active distribute list timer. Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
2010-04-18ospfd: Make sure all external routes are updated.Joakim Tjernlund
Roman Hoog Antink <rha@open.ch> reports: When adding a connected route (using vtysh, without restart) to the redistribution access list of ospfd, while static routes already exist, the update timer ospf_distribute_list_update_timer() is being run for static routes only. That way, the connected route never appears in the OSPF database, until quagga is completely restarted. The update timer for connected routes is cancelled in ospfd/ospfd_zebra.c:ospf_distribute_list_update():976, were a new timer is scheduled for static routes, caused by the loop in ospf_filter_update(). * ospf_zebra.c: (ospf_distribute_list_update_timer) make it refresh all external routes. This fixes the problem reported by Roman. Signed-off-by: Joakim Tjernlund <Joakim.Tjernlund@transmode.se>
2010-04-18git: add pointers to out-of-tree workGreg Troxel
* (HACKING.pending): List known places where work on quagga that is not (yet) integrated into the main tree can be found.
2010-04-14git: add (generated) m4 files to .gitignoreGreg Troxel
2010-03-11doc: fixed spelling in bgpd.texiIvan Moskalyov
2010-03-10build: Add QuaggaId to README.NetBSDGreg Troxel
2010-03-09release: 0.99.16Denis Ovsienko
* configure.ac: bump up version number
2010-01-27lib: undo the last PIE commitStephen Hemminger
Stephen: A recent change breaks build Quagga on Debian Lenny with amd64 (but not i386). Not sure whether this is cause by ld, gcc, or libtool; but whatever it just won't work for a large number of users. Mathias: I would guess it's a problem of libtool because it adds "-fPIC -DPIC" to the compiler flags but doesn't remove -fPIE. That wouldn't be a problem if the compiler would ignore the -fPIE in favor to the latter -fPIC, but obviously it does not. [...] Those objects were actually compiled with -fPIC but also -fPIE which seems to confuse gcc.