Merge pull request #20322 from opensourcerouting/fix/treat-as-withdraw-if-lenghts-are-conflictingbgpd: Treat as withdraw iBGP session when total attr length is path attributes
zebra: [TP] Traces in Zebra netlink parse error conditionsAdding traces in Zebra netlink parse error conditions
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
zebra: [TP] Traces in Zebra EVPN operationsAdding traces in Zebra EVPN operations
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
zebra: [TP] Traces in VTEP operationAdding traces in Zebra VTEP operation
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
zebra: [TP] Traces in NHG operationsAdding traces in Zebra NHG operations
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
zebra: [TP] Traces in Route or Rib operationsAdding traces in Zebra Route/Rib operations
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
lib: Add Nexthop status in nexthop2str functionAdd Active/Inactive Nexthop status in nexthop2str function
Ticket :# 4265339
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
zebra: [TP] Traces in Intf managementAdding traces in Zebra Intf management
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
bgpd: Treat as withdraw iBGP session when total attr length is path attributesRFC 7606 says:
Attribute Length Fields
There are two error cases in which the Total Attribute Length value
can be in conflict with the enclosed path attributes, which
themselves carry length values:
* In the first case, the length of the last encountered path
attribute would cause the Total Attribute Length to be exceeded
when parsing the enclosed path attributes.
...
tests: Add a test_ospfv3_route_map_forwarding testThe route-map set command `set forwarding-address X::X`
in ospf6d has a bug where this was not correctly being
honored. Let's write a topotest that shows that this
fix works and is correct.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
ospf6d: Route-Map parameter forwarding-address functionality not working properlyIssue:
The forwarding-address parameter set via a route-map is not being reflected
in the OSPFv3 LSAs as expected.
Specifically, after configuring the following on router L1,
the forwarding-address is missing from the OSPFv3 database
on the peer router S1:
Root Cause:
In the function ospf6_asbr_redistribute_add,
the forwarding address is initially set from the route-map:
if (!IN6_IS_ADDR_UNSPE...
bgpd: Allow LL peering to update v6 GUACurrently, when a LL based peer in BGP is created:
neighbor fe80:1::1 remote-as external
neighbor fe80:1::1 interface r2-eth0
When the interface is updated with a v6 GUA after
peering is established. The v6 GUA is never sent to the
peer.
Current code behavior allows for bgp unnumbered peering:
neighbor r2-eth1 remote-as external
To update it's GUA address when it comes the address i...
Merge pull request #20298 from opensourcerouting/fix/staticd_delete_route_blackhole_typestaticd: Prevent deleting a static route if blackhole type is not the same
zebra: add CLI 'no' versions for max-bw and othersAdd '[no]' versions for commands:
* max-bw
* max-rsv-bw
* unrsv-bw
Without these frr-reload failed after deletion of link-params with these commands.
For them to work update:
* lib_interface_zebra_link_params_max_bandwidth_destroy
* lib_interface_zebra_link_params_max_reservable_bandwidth_destroy
* lib_interface_zebra_link_params_unreserved_bandwidths_unreserved_bandwidth_destroy
All three...
ripd: Speed up convergence of rip_default_metricThis test is changing the default metric and then waiting
30 seconds to ensure that the value is passed. This is a
bit problematic in that rip operates on a 30 second window
before updates are sent. Let's change the basic update
operation in rip to 5 seconds and see if the test will
not occassionally fail now.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
ospf6d: Fix handling of default-routesWhen using ZEBRA_ROUTE_MAX as upper limit for loops, we have to remember
that DEFAULT_ROUTE == ZEBRA_ROUTE_MAX and we must therefore also include this
last type.
Signed-off-by: Alexander Rose <alexander.rose@secunet.com>
bgpd: [TP] Traces in the BGP Update processingAdding new tracepoints in the BGP Update processing
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
bgpd: [TP] Traces in the BGP UpdgrpAdding new tracepoints in the BGP Updgrp
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
bgpd: [TP] Traces in the BGP-Zebra interactionAdding new tracepoints in the BGP-Zebra interactions
co-authored by Nvidia FRR team
Signed-off-by: Rajasekar Raja <rajasekarr@nvidia.com>
bgpd: Fix Coverity analysisThis fixes coverity warning: Assigning value "NULL" to "pfx"
here, but that stored value is overwritten before it can be used
Removed unnessary NULL assignement to const struct prefix *pfx
- The line pfx = NULL; assigns NULL to the variable
- Immediately after, the very next statement checks a condition
and overwrites pfx with a different value
- The NULL value is never actually used - ...
zebra: don't access ifp from dplane pthreadThe function that handles incoming IPNEIGH netlink messages
runs in the dplane pthread - it can't look up vrfs or ifps.
Signed-off-by: Mark Stapp <mjs@cisco.com>
Donald Sharp
Rajasekar Raja
Donatas Abraitis
Donald Sharp
Kyrylo Yatsenko
Alexander Rose
Mark Stapp