msg_new takes a uint16_t, the length passed
down variable is a unsigned int, thus 32 bit.
It's possible, but highly unlikely, that the
msglen could be greater than 16 bit.
Let's just add some checks to ensure that
this could not happen.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
(cherry picked from commit 283cc51178)
Duplicate next-hops are maintained for OSPF inter-area and AS
external routes in the OSPF routing table as long as they
correspond to LSAs for different adverting routers. The
intra-area route computation will not result in duplicate
next-hops.
Signed-off-by: Acee Lindem <acee@lindem.com>
Since the timer thread for ```OSPF_ROUTE_AGGR_DEL``` has been created,
the subsequent "no summary-address" commands shouldn't trigger redundant timers.
Signed-off-by: anlan_cs <anlan_cs@126.com>
The users would know the failure of some case for the two commands, just
return the error ```CMD_WARNING_CONFIG_FAILED```.
Signed-off-by: anlan_cs <anlan_cs@126.com>
When OSPFd starts, there is 2 possible scenarios for Segment Routing:
1/ Routes associated to Prefixes are not yet available i.e. Segment Routing LSA
are received before LSA Type 1. In this case, the function
ospf_sr_nhlfe_update() is triggered when a new SPF is launch. Thus, neighbors
and output label are always synchronise with the routing table.
2/ Routes are already available i.e. LSA Type 1 are received before Segment
Routing LSA, in particular the Router Information which contains the SRGB.
During nhlfe computation, perfixes are leave with incomplete configuration, in
particular, the SR nexthop is set to NULL. If this scenario is handle through
the function update_out_nhlfe (triggered when SRGB is received or modified from
a neighbor node), the output label is not correctly configured as the nexthop
SR node associated to the prefix has been leave to NULL.
This patch correct this problem by calling the function compute_nhlfe() when
the nexthop SR Node associated to the prefix is NULL within the
update_out_nhlfe() function. Thus, we guarantee that the SR prefix is always
correctly configuration indpedently of the scenario i.e. arrival of the
different LSA.
Signed-off-by: Olivier Dugeon <olivier.dugeon@orange.com>
When originating a default AS-External LSA in one OSPF instance,
it wasn't working if the criteria route was installed by another OSPF
instance. This required more flexible processing of the OSPF external
route information.
Also fix problem multi-instance display for "show ip ospf
<instance> database ...".
Signed-off-by: Acee Lindem <acee@lindem.com>
Under certain timing conditions, the current logic asserts in
ospf_lsa_refresh_delay(). While this isn't readily reproducible,
the only explanation is that the conversion from struct timeval
to milliseconds is 0 due to rounding off the microseconds.
Signed-off-by: Acee <aceelindem@gmail.com>
If we set router-id, e.g. `router-id x.x.x.x`, then we have:
```
pc.donatas.net# show ip ospf | include Router ID
OSPF Routing Process, Router ID: x.x.x.x
```
But once we remove it (`no router-id x.x.x.x`), the old router-id remains.
This is kinda OK, but to be consistent with OSPFv3 we should use what Zebra
already has, instead of retaining the old one.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
A new command is `ospf router-id ...`, but the old one is also valid. Just a no
form was missed.
Signed-off-by: Donatas Abraitis <donatas@opensourcerouting.org>
The configured OSPF refresh interval was not being used for opaque LSA (it always used the constant). Also, modified the timers lsa min-arrival command to have a maximum of 5000 msecs as well as providing a path for backward command compatibility.
Added missing user documentation for both timers lsa min-arrival and timers throttle lsa all.
Signed-off-by: Acee Lindem <acee@lindem.com>
Fix syntax of the following no commands:
1. `no area virtual link A.B.C.D hello-interval <NUM>`, `<NUM>` can be omitted.
2. `no area nssa default-information-originate metric <NUM>`, `<NUM>` can be omitted.
3. `no area nssa range cost <NUM>`, `<NUM>` can be omitted.
4. `no area default cost <NUM>`, `<NUM>` can be omitted.
5. `no ospf write-multiplier <NUM>`, `<NUM>` can be omitted.
6. `no default-information originate metric <NUM>`, `<NUM>` can be omitted.
7. `no distance <NUM>`, `<NUM>` can be omitted.
Signed-off-by: Shbinging <bingshui@smail.nju.edu.cn>
When we use the no area X.X.X.X range A.B.C.D/M command, if the area no longer has an interface to which it belongs, then the area should be deleted from the LSDB. This processing logic is consistent with instructions such as no network area and no area authentication.
Signed-off-by: Shbinging <bingshui@smail.nju.edu.cn>
In the processing of nssa, if the number of areas that need to be translated is greater than 0, then abr will be regarded as asbr, and it will be marked (0x3) in the flag of router lsa. When a certain area is set from nssa to a normal area, the areas that need to be translated may be reduced. The asbr should be re-interpreted as abr when the translated area is 0.
Signed-off-by: Shbinging <bingshui@smail.nju.edu.cn>
Our standard says when in a .h file and declaring a function all parameters
must have their variables spelled out. Let's do this for ospf_lsa.h
Modified ospf_lsa.c to also use event instead of thread, and uint8_t instead
of unsigned short.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
When we use the no area X.X.X.X range A.B.C.D/M command, if the area no longer has an interface to which it belongs, then the area should be deleted from the LSDB. This processing logic is consistent with instructions such as no network area and no area authentication.
Signed-off-by: Shbinging <bingshui@smail.nju.edu.cn>
When 'ip ospf neighbor-filter <filter-name>' is configured in the startup
configuration, it is in the running configuration but not applied on the
interface.
Signed-off-by: Acee Lindem <acee@lindem.com>
When parsing the SR-Algorithm TLV in the OSPF Router Information Opaque
LSA, assure that not more than the maximum number of supported
algorithms are copied from the TLV.
Signed-off-by: Acee Lindem <acee@lindem.com>
This file had a bunch of debugs that were operating on
specific vrf's but not telling you which vrf this was
happening in. Add code to allow operator debugging
to understand what vrf is being worked on.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
When "ip ospf network point-to-multipoint delayed-reflooding" is configured,
LSAs received on an OSPF P2MP network are not reflooded. Since LSA reflooding
would normally serve as an implied LSA acknowledgment, an explicit OSPF ack
should be sent to avoid retransmission by the neighbor which orginally flooded
the LSA on the P2MP network.
Signed-off-by: Acee Lindem <acee@lindem.com>
When an ospf interface is not in the backbone area, but it receives a
packet from the backbone, no logs are generated for this mismatch.
However, the opposite scenario does generate logs.
Add a log for this case.
Signed-off-by: Loïc Sang <loic.sang@6wind.com>
When enabling "mpls ldp-sync" under "router ospf" ospfd configures
SET_FLAG(ldp_sync_info->flags, LDP_SYNC_FLAG_IF_CONFIG) so internally knowing
that the ldp-sync feature is enabled. However the flag is not cleared when
turning of the feature using "nompls ldp-sync"!
https://github.com/FRRouting/frr/issues/16375
Signed-off-by: Christian Breunig <christian@breunig.cc>
In the "2x2 matrix" of these, I accidentally edited "row-wise" when I
should've edited "column-wise"... *sigh*
Reported-by: github.com/rbfnet
Fixes: #16349
Fixes: 110945ba0d ("ospfd: fix GR state location")
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
1. On P2MP interfaces, direct ack would include the same LSA multiple times
multiple packets are processed before the OSPF interfae direct LSA
acknowledgment event is processed. Now duplicates LSA in the same event
are suppressed.
2. On non-broadcast interfaces, direct acks for multiple neighbors would be
unicast to the same neighbor due to the multiple OSPF LS Update packets
being process prior to the OSPF interface direct ack event. Now, separate
direct acks are unicast to the neighbors requiring them.
3. The interface delayed acknowledgment timer runs would run continously
(every second as long as the interace is up). Now, the timer is set
when delayed acknowledgments are queued and all queued delayed
acknowledges are sent when it fires.
4. For non-broadcast interface delayed acknowledgments, the logic to send
to multiple neighbors wasn't working because the list was emptied while
building the packet for the first neighbor.
Signed-off-by: Acee Lindem <acee@lindem.com>
The current OSPF neighbor retransmission operates on a single per-neighbor
periodic timer that sends all LSAs on the list when it expires.
Additionally, since it skips the first retransmission of received LSAs so
that at least the retransmission interval (resulting in a delay of between
the retransmission interval and twice the interval. In environments where
the links are lossy on P2MP networks with "delay-reflood" configured (which
relies on neighbor retransmission in partial meshs), the implementation
is sub-optimal (to say the least).
This commit reimplements OSPF neighbor retransmission as follows:
1. A new data structure making use the application managed
typesafe.h doubly linked list implements an OSPF LSA
list where each node includes a timestamp.
2. The existing neighbor LS retransmission LSDB data structure
is augmented with a pointer to the list node on the LSA
list to faciliate O(1) removal when the LSA is acknowledged.
3. The neighbor LS retransmission timer is set to the expiration
timer of the LSA at the top of the list.
4. When the timer expires, LSAs are retransmitted that within
the window of the current time and a small delta (50 milli-secs
default). The LSAs that are retransmited are given an updated
retransmission time and moved to the end of the LSA list.
5. Configuration is added to set the "retransmission-window" to a
value other than 50 milliseconds.
6. Neighbor and interface LSA retransmission counters are added
to provide insight into the lossiness of the links. However,
these will increment quickly on non-fully meshed P2MP networks
with "delay-reflood" configured.
7. Added a topotest to exercise the implementation on a non-fully
meshed P2MP network with "delay-reflood" configured. The
alternative was to use existing mechanisms to instroduce loss
but these seem less determistic in a topotest.
Signed-off-by: Acee Lindem <acee@lindem.com>
