In practical terms, unplanned GR refers to the act of recovering
from a software crash without affecting the forwarding plane.
Unplanned GR and Planned GR work virtually the same, except for the
following difference: on planned GR, the router sends the Grace-LSAs
*before* restarting, whereas in unplanned GR the router sends the
Grace-LSAs immediately *after* restarting.
For unplanned GR to work, ospf6d was modified to send a
ZEBRA_CLIENT_GR_CAPABILITIES message to zebra as soon as GR is
enabled. This causes zebra to freeze the OSPF routes in the RIB as
soon as the ospfd daemon dies, for as long as the configured grace
period (the defaults is 120 seconds). Similarly, ospfd now stores in
non-volatile memory that GR is enabled as soon as GR is configured.
Those two things are no longer done during the GR preparation phase,
which only happens for planned GRs.
Unplanned GR will only take effect when the daemon is killed
abruptly (e.g. SIGSEGV, SIGKILL), otherwise all OSPF routes will
be uninstalled while ospfd is exiting. Once ospfd starts, it will
check whether GR is enabled and enter in the GR mode if necessary,
sending Grace-LSAs out all operational interfaces.
One disadvantage of unplanned GR is that the neighboring routers
might time out their corresponding adjacencies if ospfd takes too
long to come back up. This is especially the case when short dead
intervals are used (or BFD). For this and other reasons, planned
GR should be preferred whenever possible.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
When using route maps with external routes in OSPF as follows:
```
set metric +10
```
The current behavior is to use the default ospf metric as the base and then add
to 10 to it. The behavior isn't useful as-is. A value 30 (20 dfeault + 10) can
be set directly instead. the behavior is also not consistent with bgp. bgp does
use the rib metric in this case as the base. The current behavior also doesn't
allow the metric to accumulate when crossing different routing domains such as
vrfs causing the metric to reset every time the route enters a new vrf with a new
ospf network.
This PR changes the behavior such that the rib metric is used as a base for
ospf exteral routes when used with `set metric -/+`
Signed-off-by: Jafar Al-Gharaibeh <jafar@atcorp.com>
Add the "default-information-originate" option to the "area X nssa"
command. That option allows the origination of Type-7 default routes
on NSSA ABRs and ASBRs.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Effectively a massive search and replace of
`struct thread` to `struct event`. Using the
term `thread` gives people the thought that
this event system is a pthread when it is not
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
This is a first in a series of commits, whose goal is to rename
the thread system in FRR to an event system. There is a continual
problem where people are confusing `struct thread` with a true
pthread. In reality, our entire thread.c is an event system.
In this commit rename the thread.[ch] files to event.[ch].
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Fix obvious bug where the wrong area filter-lists were being updated
in response to a prefix-list update.
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
That commit aim is to fix an invalid behavior when
default-information is activated on ospf router without always option.
Consider an ASBR with:
-one default route coming from ospf,
-and another default route coming from another deaemon (such BGP or static).
When the daemon bgp stops advertising its default route,
-ospf continues to advertise its previous default route (with aging 0),
-this may create default routing loops.
Expected behavior: is to update the removed external default route with
MAXAGING value.
Updating with MAXAGING value will notify the fact the route is currently
invalid. A later removal from ospf external LSA database will be made.
Analysis: all default routes have their type overwritten by a
DEFAULT_ROUTE type. Thus all default routes whatever its origin (ospf,
bgp, static...) is treated in a same way. But this is not pertinent for
ospf originated default routes.
Fix: avoid overwiting of route type when default route is ospf type.
Signed-off-by: Francois Dumontet <francois.dumontet@6wind.com>
Update ospfd and ospf6d to send opaque route attributes to
zebra. Those attributes are stored in the RIB and can be viewed
using the "show ip[v6] route" commands (other than that, they are
completely ignored by zebra).
Example:
```
debian# show ip route 192.168.1.0/24
Routing entry for 192.168.1.0/24
Known via "ospf", distance 110, metric 20, best
Last update 01:57:08 ago
* 10.0.1.2, via eth-rt2, weight 1
OSPF path type : External-2
OSPF tag : 0
debian#
debian# show ip route 192.168.1.0/24 json
{
"192.168.1.0\/24":[
{
"prefix":"192.168.1.0\/24",
"prefixLen":24,
"protocol":"ospf",
"vrfId":0,
"vrfName":"default",
"selected":true,
[snip]
"ospfPathType":"External-2",
"ospfTag":"0"
}
]
}
```
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
This removes a giant `switch { }` block from lib/zclient.c and
harmonizes all zclient callback function types to be the same (some had
a subset of the args, some had a void return, now they all have
ZAPI_CALLBACK_ARGS and int return.)
Apart from getting rid of the giant switch, this is a minor security
benefit since the function pointers are now in a `const` array, so they
can't be overwritten by e.g. heap overflows for code execution anymore.
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
ospf_distribute_list_update currently passes two arguments to
ospf_distribute_list_update_timer - pointer to the ospf structure and
protocol type. The protocol type is only used for logging and is not
even correct because if multiple changes happen during one
ospf->min_ls_interval, then only the type of the first change is logged.
It is better to completely remove the protocol type argument to have a
correct log and eliminate the need for memory allocation.
Signed-off-by: Igor Ryzhov <iryzhov@nfware.com>
Move `is_default_prefix` variations to `lib/prefix.h` and make the code
use the library version instead of implementing it again.
NOTE
----
The function was split into per family versions to cover all types.
Using `union prefixconstptr` is not possible due to static analyzer
warnings which cause CI to fail.
The specific cases that would cause this failure were:
- Caller used `struct prefix_ipv4` and called the generic function.
- `is_default_prefix` with signature using `const struct prefix *` or
`union prefixconstptr`.
The compiler would complain about reading bytes outside of the memory
bounds even though it did not take into account the `prefix->family`
part.
Signed-off-by: Rafael Zalamena <rzalamena@opensourcerouting.org>
RFC 3623 specifies the Graceful Restart enhancement to the OSPF
routing protocol. This PR implements support for the restarting mode,
whereas the helper mode was implemented by #6811.
This work is based on #6782, which implemented the pre-restart part
and settled the foundations for the post-restart part (behavioral
changes, GR exit conditions, and on-exit actions).
Here's a quick summary of how the GR restarting mode works:
* GR can be enabled on a per-instance basis using the `graceful-restart
[grace-period (1-1800)]` command;
* To perform a graceful shutdown, the `graceful-restart prepare ospf`
EXEC-level command needs to be issued before restarting the ospfd
daemon (there's no specific requirement on how the daemon should
be restarted);
* `graceful-restart prepare ospf` will initiate the graceful restart
for all GR-enabled instances by taking the following actions:
o Flooding Grace-LSAs over all interfaces
o Freezing the OSPF routes in the RIB
o Saving the end of the grace period in non-volatile memory (a JSON
file stored in `$frr_statedir`)
* Once ospfd is started again, it will follow the procedures
described in RFC 3623 until it detects it's time to exit the graceful
restart (either successfully or unsuccessfully).
Testing done:
* New topotest featuring a multi-area OSPF topology (including stub
and NSSA areas);
* Successful interop tests against IOS-XR routers acting as helpers.
Co-authored-by: GalaxyGorilla <sascha@netdef.org>
Signed-off-by: Renato Westphal <renato@opensourcerouting.org>
Adding defensive code to the interface_link_params zebra callback
to check if the link params changed before taking action.
Signed-off-by: Karen Schoener <karen@voltanet.io>
Currently ospfd relies on vrf bitmaps in zclient to check that the
redistribution is configured. This doesn't work when the VRF for OSPF
instance doesn't exist yet, because vrf bitmaps ignore VRF_UNKNOWN id.
Because of this, the following problems occur when the VRF doesn't exist:
- repeated "redistribute smth" command is processed as a first-time
instead of an update
- "no redistribute smth" doesn't work at all
This commit fixes both issues by relying on internal redistribution
config instead of zclient vrf bitmaps.
Signed-off-by: Igor Ryzhov <iryzhov@nfware.com>
Found a couple spots where FRR was using `case default` when
using a switch over an enum. In this case we *must* enumerate
all states as part of the switch.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Description:
OSPF does not have an option to control the maximum multiple
equal cost paths to reach a destination/route(ECMP).
Currently, it is using the system specific max multiple paths.
But Somtimes, It requires to control the multiple paths from ospf.
This cli helps to configure the max number multiple paths in ospf.
Signed-off-by: Rajesh Girada <rgirada@vmware.com>
This patch allows to store Link State Information received through the various
LSAs into a dedicated Traffic Engineering Database (TED). This feature is
automatically activated once mpls-te is enabled.
A new CLI command `mpls-te export` permits to export the TED to other daemons
through the new ZAPI Opaque Link State messages. In complement, a new CLI
command `show ip ospf mpls-te database ...` output the contains of the TED to
the console.
Major modifications take place in ospf_te.[c, h]. File ospf_zebra.c has been
modified to handle TED synchronisation request.
Signed-off-by: Olivier Dugeon <olivier.dugeon@orange.com>
Back when I put this together in 2015, ISO C11 was still reasonably new
and we couldn't require it just yet. Without ISO C11, there is no
"good" way (only bad hacks) to require a semicolon after a macro that
ends with a function definition. And if you added one anyway, you'd get
"spurious semicolon" warnings on some compilers...
With C11, `_Static_assert()` at the end of a macro will make it so that
the semicolon is properly required, consumed, and not warned about.
Consistently requiring semicolons after "file-level" macros matches
Linux kernel coding style and helps some editors against mis-syntax'ing
these macros.
Signed-off-by: David Lamparter <equinox@diac24.net>
This patch solves 2 Segment Routing prefix bugs:
- If Segment Routing is not enabled in the initial configuration, Extended
Prefix Opaque LSA is not flood. This is due to a control flag which is
set only when Segment Routing is enabled at startup and not latter.
- Attempting to modify Segment Routing prefix flag e.g. adding or removing
no-php or explicit-null flag, doesn't work as expected: Corresponding entry
in the MPLS table is not updated, Extended Prefix Opaque LSA carry wrong flag
value, and neighbor set a wrong configuration in the MPLS table for this
Segment Routing prefix.
The first bug is corrected in ospfd/ospf_ext.c:
- Flag setting is moved from ospf_ext_ism_change() to set_ext_prefix() function
The seconf one is corrected in ospfd/ospf_sr.c:
- For self node, previous MPLS entry is removed if needed and flag reset before
setting the new Segment Routing prefix configuration
- For neighbor node, srnext field of sr_prefix structure is always set and not
only for new SR Prefix.
Signed-off-by: Olivier Dugeon <olivier.dugeon@orange.com>
Issue #7926 hilight a race condition in Segment Routing processing.
The problem occurs when Router Information Opaque LSA is received late, in
particular after SPF run and after ospf_sr_nhlfe_update() was called. This
scenario is unfrequent and takes place due to a slow DR election.
In this particular case, SR Prefix are handle but not fully fill. In fact,
SRGB for the nexthop is not yet received and thus, output label could not
be computed.
When Router Information Opaque LSA is received and processed, if the
corresponding SR node is a direct neighbor of the self node, update_out_nhlfe()
is called against all SR nodes to adjust SR prefix if the next hop is the new
SR node. The function wrongly computes output label and configure a bad MPLS
LFIB entries.
Another way to hilight the problem is to change through CLI the SRGB of a node
and look to MPLS LFIB of direct neighbor, in particular those who announce
EXPLICIT NULL Prefix SID.
This patch correct the update_out_nhlfe() function by calling the appropriate
function (sr_prefix_out_label() instead of index2label()) to compute the output
label.
Some log debugs were adjusted and unused prefix route table was removed too.
Signed-off-by: Olivier Dugeon <olivier.dugeon@orange.com>
Removing the obsolete ldp-sync periodic 'hello' message.
When ldp-sync is configured, IGPs take action if the LDP process goes down.
The IGPs have been updated to use the zapi client close callback to detect
the LDP process going down.
Signed-off-by: Karen Schoener <karen@voltanet.io>
When ldp-sync is configured, IGPs take action if the LDP process goes down.
Currently, IGPs detect the LDP process is down if they do not receive a
periodic 'hello' message from LDP within 1 second.
Intermittently, this heartbeat mechanism causes false topotest failures.
When the failure occurs, LDP is busy receiving messages from zebra for a
few seconds. During this time, LDP does not send the expected periodic
message.
With this change, IGPs detect LDP down via zapi client close message.
Signed-off-by: Karen Schoener <karen@voltanet.io>
The `enum zclient_send_status` enum needs to be extended
throughout the code base to use the new states and
to fix up places where we tested against the return
value being non zero.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
The route_map_object_t was being used to track what protocol we were
being called against. But each protocol was only ever calling itself.
So we had a variable that was only ever being passed in from route_map_apply
that had to be carried against and everyone was testing if that variable
was for their own stack.
Clean up this route_map_object_t from the entire system. We should
speed some stuff up. Yes I know not a bunch but this will add up.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Description:
Apis for creating/deleting aggregate routes.
Origination of summary route on behalf of matched external routes.
Signed-off-by: Rajesh Girada <rgirada@vmware.com>
