Stephen Worley [Tue, 30 Apr 2019 14:43:16 +0000 (10:43 -0400)]
zebra: Check on startup route_info has all types
Add a function to check if the route_info array
has all types specified with data in it. Specifically,
test the 'key' attribute for non-zero data. Ignore
ZEBRA_ROUTE_SYSTEM as it should be zero key anyway.
Signed-off-by: Stephen Worley <sworley@cumulusnetworks.com>
Donald Sharp [Wed, 1 May 2019 01:29:03 +0000 (21:29 -0400)]
zebra: Fix incorrect reading of REMOTE_VTEP_[ADD|DEL]
With flooding control added recently we were not properly handling
the new flood control parameter in zebra_vxlan.c handler functions.
The error message that was being repeatedly seen:
2019/05/01 00:47:32 ZEBRA: [EC 100663311] stream_get2: Attempt to get out of bounds
2019/05/01 00:47:32 ZEBRA: [EC 100663311] &(struct stream): 0x7f0f04001740, size: 22, getp: 22, endp: 22
The fix was to ensure that both the _add and _del functions kept proper
sizing of amount of data read *and* the _del function was not
reading the flood_control data from the stream.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Link-ID and Remote IP address must be set accordingly to the interface type
(Point-to-Point or Broadcast) from the neighbor information. However, this
information are only valid once the Network State Machine (NSM) is Full i.e.
when the adjacency is up. The original TE code only look to Interface State
Machine (ISM) change which not allow to collect valid neighbor information.
The patch move setup of Link-ID and Remote-IP TE parameters from
ospf_mpls_te_ism_change() to ospf_mpls_te_nsm_change() function.
David Lamparter [Mon, 4 Feb 2019 00:22:03 +0000 (01:22 +0100)]
isisd: replace dict_* with DECLARE_RBTREE
Historically, isisd has been carrying around its own red-black tree to
manage its LSP DB in. This replaces that with the newly-added
DECLARE_RBTREE_*. This allows completely removing the dict_* code.
Signed-off-by: David Lamparter <equinox@diac24.net>
David Lamparter [Thu, 31 Jan 2019 01:12:38 +0000 (02:12 +0100)]
lib: use DECLARE_LIST for thread_list
Replaces the open-coded thread_list with a DECLARE_LIST instantiation.
Some function prototypes are actually identical to what was previously
open-coded.
Signed-off-by: David Lamparter <equinox@diac24.net>
David Lamparter [Mon, 18 Feb 2019 20:17:22 +0000 (21:17 +0100)]
tests: exercise the typesafe list wrappers
Since all of these list implementations provide almost the same API, we
can run and validate them against the same test code. 9 tests for the
price of one!
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
David Lamparter [Thu, 31 Jan 2019 00:09:13 +0000 (01:09 +0100)]
lib: typesafe lists, skiplist & hash
By the power of the C preprocessor, these macros provide type-safe
warppers for simple lists, skiplists and hash tables. Also, by changing
the instantiation macro, it is easily possible to switch between
algorithms; the code itself does not need to be changed since the API
is identical across all algorithms.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
lib: introduce a read-write lock for northbound configurations
The upcoming gRPC-based northbound plugin will run on a separate
pthread, and it will need to have access to the running configuration
global variable. Introduce a rw-lock to control concurrent access
to the running configuration. Add the lock inside the "nb_config"
structure so that it can be used to protect candidate configurations
as well (this might be necessary depending on the threading scheme
of future northbound plugins).
Renato Westphal [Sat, 2 Feb 2019 20:03:15 +0000 (18:03 -0200)]
lib: add API to allow northbound clients to lock/unlock the running configuration
The ability to lock the running configuration to prevent other users
from changing it is a very important one. We already supported
the "configure exclusive" command but the lock was applied to
the CLI users only (other clients like ConfD could still commit
configuration transactions, ignoring the CLI lock). This commit
introduces a global lock for the running configuration that is
shared by all northbound clients, and provides a public API to
manipulate it. This way other northbound clients will also be able
to lock/unlock the running configuration if required (the upcoming
gRPC northbound plugin will have RPCs for that).
NOTE: this is a management-level lock for the running configuration,
not to be confused with low-level locks used to avoid data races.
Don Slice [Wed, 24 Apr 2019 17:14:57 +0000 (17:14 +0000)]
zebra: resolve issue with protocol route-map not applied properly
Problem reported that route-maps applied to "ip protocol table bgp"
would not be invoked if the ip protocol table command was issued
after the bgp prefixes were installed. Found that a recent change
improving how often nexthop_active_update runs missed causing this
filtering to be applied. This fix resolves that issue as well as
a couple of other places that were problematic with the recent
change.
Signed-off-by: Don Slice <dslice@cumulusnetworks.com>
Don Slice [Tue, 23 Apr 2019 14:54:56 +0000 (10:54 -0400)]
tools: frr-reload.py stop bouncing peers on bfd timer change
Problem reported that if a bgp neighbor had a bfd timer change
made in frr.conf and systemctl reload frr performed, the neighbor
with the timer changed bounced. If the change is made in vtysh
by just adding the new timer values, no peer bounce occurs. This
fix skips the delete part of the delete/add process in frr-reload
so the peers stay up.
Signed-off-by: Don Slice <dslice@cumulusnetworks.com>
- Change MPLS-TE from global to per Area
- Add new mpls_te_area structure to area in replacement of global variable
isisMPLS_TE
- Move mpls-te from global to instance in frr-isisd.yang
- Change code in isis_te.c, isis_northbound.c, isis_cli.c, isis_pdu.c,
isis_lsp.c and isis_zebra.c accordingly
Donald Sharp [Sat, 27 Apr 2019 23:55:21 +0000 (19:55 -0400)]
zebra: Modify how we display/store os description
The alias/description of an interface in linux was being
used to override the internal description. As such let's
fix the display to keep track of both if we have it.
Config in FRR:
!
interface docker0
description another combination
!
interface enp3s0
description BAMBOOZLE ME WILL YOU
!
Config in linux:
sharpd@robot ~/f/zebra> ip link show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 65536 qdisc noqueue state UNKNOWN mode DEFAULT group default qlen 1000
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
alias This is the loopback you cabbage
2: enp3s0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc fq_codel state UP mode DEFAULT group default qlen 1000
link/ether 74:d0:2b:9c:16:eb brd ff:ff:ff:ff:ff:ff
alias HI HI HI
Now the 'show int descr' command:
robot# show int description
Interface Status Protocol Description
docker0 up down another combination
enp3s0 up up BAMBOOZLE ME WILL YOU
HI HI HI
lo up up This is the loopback you cabbage
Fixes: #4191 Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Quentin Young [Wed, 24 Apr 2019 17:23:12 +0000 (17:23 +0000)]
doc: add copyright notice from Ralph Keller
Ralph has kindly granted us GPLv2+ license to use this documentation,
and requests that we keep a reference to his name. Add these facts to
the documentation.
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
bgpd: Prevent IPv6 routes received via a ibgp session with own ip as nexthop
Prevent IPv6 routes received via a ibgp session with one of its own interface
ip as nexthop from getting installed in the BGP table.
Implemented IPV6 HASH table, where we need to add any ipv6 address as they
gets configured and delete them from the HASH table as the ipv6 addresses
get unconfigured. The above hash table is used to verify if any route learned
via BGP has nexthop which is equal to one of its its connected ipv6 interface.
Donald Sharp [Wed, 24 Apr 2019 02:01:44 +0000 (22:01 -0400)]
pimd: Add missing flags to json output
the json code has not been updated since a variety of new flags have
been added to the code base. Add those flags in so we can tell
what is going on sometimes.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Donald Sharp [Mon, 22 Apr 2019 21:36:58 +0000 (17:36 -0400)]
pimd: When creating new upstream state, figure out what we should join
Always when creating a new S,G state look at all possible ifchannels
to decide what the mroute should be.
The bug that this is fixing is this:
Suppose two incoming `*,G` joins on swp1, and swp2.
Now suppose that one of those ifchannel `*,G` sends a `*,G S,G RPT Prune`.
We were creating the S,G upstream state as we should but we were
only looking at the S,G ifchannel to decide the S,G mroute we would
be creating. As such what we need to do is to look over the associated
*,G ifchannels and allow us to associate correct oil needed.
Ticket: CM-24732 Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Quentin Young [Mon, 22 Apr 2019 20:22:00 +0000 (20:22 +0000)]
doc: cleanup OSPF API server documentation
* Reflow to 80 columns
* Improve markup
* Add --apiserver option to example ospfd invocations
* Add note on requirement of this option to use api server
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
projects / mirror_frr.git / log