RCA: When upstream transition from Joined to NotJoined due to SGRpt
prune, then only SGRpt prune was sent and SG Prune is missed.
Fix: Send SG Prune towards source as well as SGRpt prune towards RP.
Signed-off-by: sarita patra <saritap@vmware.com>
The Bootstrap message RX path needs a RPF check for the BSR address,
and this is implemented both incorrectly as well as quite ugly.
Clean up and fix case when we have multiple interfaces to the same LAN
and/or ECMP nexthops (both would cause message duplication, the former
can even cause BSM forwarding loops.)
Signed-off-by: David Lamparter <equinox@opensourcerouting.org>
enum based switches should never use default. It makes
it very hard to fix and find issues when the enum is
changed.
Signed-off-by: Donald Sharp <sharpd@nvidia.com>
Problem Statement:
==================
Mroutes are not recovered after shut/no shut of DUT to RP links
One interface is not added in OIL List in intermediate router,
hence traffic never received at LHR and mroutes not created for (S,G).
Root Cause Analysis:
====================
Generally (*,G) PIM Join is received first and then (S,G) joins are received.
This issue occurs when (S,G) join comes first and then the (*,G) Join.
When (S,G) PIM Join is received, ifchannel is created and channel_oil
OIF flag is set to PIM_OIF_FLAG_PROTO_PIM. Now when (*,G) join is received
the flag PIM_OIF_FLAG_PROTO_STAR is not inherited due to wrong check present in
function pim_upstream_inherited_olist_decide.
Fix:
===================
When (*,G) PIM Join is received, it should always add PIM_OIF_FLAG_PROTO_STAR
flag for all the (S,G) channel oils no matter what order the (*,G) or (S,G)
is received.
Fixes: #9918
Signed-off-by: Mobashshera Rasool <mrasool@vmware.com>
The router->register_suppress_time is used to derive the
rp_keep_alive_time, but when the suppress time was changed, pim was
not recalculating the rp_keep_alive_time and left it at the old value.
This fix applies the changes when a new suppress_time is entered
(or removed.)
Signed-off-by: Don Slice <dslice@nvidia.com>
1. As per RFC 4601 Sec 4.5.7:
* JoinDesired(S,G) -> False, set SPTbit to false.
2. Change the debug type.
Signed-off-by: Mobashshera Rasool <mrasool@vmware.com>
Replace all lib/thread cancel macros, use thread_cancel()
everywhere. Only the THREAD_OFF macro and thread_cancel() api are
supported. Also adjust thread_cancel_async() to NULL caller's pointer (if
present).
Signed-off-by: Mark Stapp <mjs@voltanet.io>
Replace sprintf with snprintf where straightforward to do so.
- sprintf's into local scope buffers of known size are replaced with the
equivalent snprintf call
- snprintf's into local scope buffers of known size that use the buffer
size expression now use sizeof(buffer)
- sprintf(buf + strlen(buf), ...) replaced with snprintf() into temp
buffer followed by strlcat
Signed-off-by: Quentin Young <qlyoung@cumulusnetworks.com>
Replace all `random()` calls with a function called `frr_weak_random()`
and make it clear that it is only supposed to be used for weak random
applications.
Use the annotation described by the Coverity Scan documentation to
ignore `random()` call warnings.
Signed-off-by: Rafael Zalamena <rzalamena@opensourcerouting.org>
Problem: This happened in once in a while during testing the scenario multiple
times. When regstop timer expire and at that point if rpf interface doesn't
exist, the register state for the upstream gets struck in reg-prune state indefinitely.
This will not recover even when rpf comes back and traffic resumed because
register state is struck on prune.
RCA: Reg suppression expiry is keeping reg state unchanged when iif is absent.
Fix: When iif is absent during reg suppression expiry, treat it as couldreg
becoming false and move it NO_INFO state.
Signed-off-by: Saravanan K <saravanank@vmware.com>
RCA: Upstreams which are in register state other than noinfo, doesnt remove
register tunnel from oif after it becomes nonDR
Fix: scan upstreams with iif as the old dr and check if couldReg becomes false.
If couldreg becomes false from true, remove regiface and stop reg timer.
Do not disturb the entry. Later the entry shall be removed by kat expiry.
Signed-off-by: Saravanan K <saravanank@vmware.com>
RCA:
Either JP timer is used to send join or join timer.
We are not removing the group from jp aggregate during suppression.
So even if join timer is restarted, jp aggregate expiry during suppression
is sending join for the group.
Fix:
Remove the group from jp aggregate on the neighbor during jp suppression.
Signed-off-by: Saravanan K <saravanank@vmware.com>
RCA: Periodic join is mostly sent by nbr jp timer except for few scenarios by upstream join timer
Fix: If join timer not running, we have to use nbr jp timer to calculate
remaining time for next join.
Signed-off-by: Saravanan K <saravanank@vmware.com>
Issue: Client1------LHR-----(int-1)RP(int-2)------client2
Client2 send IGMP join for group G.
Client1 send IGMP join for group G.
verify show ip mroute in RP, will have 2 OIL.
Client2 send IGMP leave.
Verify show ip mroute in RP, will still have 2.
Root cause: When RP receives IGMP join from client2, it creates
a (s,g) channel oil and add the interface int-2 into oil list and
set the flag PIM_OIF_FLAG_PROTO_IGMP to int-2
Client1 send IGMP join, LHR will send a (*,G) join to RP. RP will
add the interface int-1 into the oil list of (s,g) channel_oil and
will set the flag PIM_OIF_FLAG_PROTO_IGMP and PIM_OIF_FLAG_PROTO_PIM
to the int-1 and set PIM_OIF_FLAG_PROTO_PIM to int-2 as well. It is
happening because of the pim_upstream_inherited_olist_decide() and
forward_on() get all the oil and update the flag wrongly.
So now when client 2 sends IGMP prune, RP will not remove the int-2
from oil list since both PIM_OIF_FLAG_PROTO_PIM & PIM_OIF_FLAG_PROTO_IGMP
are set, it just unset the flag PIM_OIF_FLAG_PROTO_IGMP.
Fix: Introduced new flags in if_channel, PIM_IF_FLAG_MASK_PROTO_PIM
& PIM_IF_FLAG_MASK_PROTO_IGMP. If a if_channel is created because of
pim join or pim (s,g,rpt) prune received, then set the flag
PIM_IF_FLAG_MASK_PROTO_PIM. If a if_channel is created becuase of IGMP
join received, then set the flag PIM_IF_FLAG_MASK_PROTO_IGMP.
When an interface needs to be added into the oil list check if
PIM_IF_FLAG_MASK_PROTO_PIM or PIM_IF_FLAG_MASK_PROTO_IGMP is set, then
update oil flag accordingly.
Signed-off-by: Sarita Patra <saritap@vmware.com>
When the (S,G) KAT expires we need to poll for activity before dropping the
entry as traffic may have been forwarded by the dataplane since the last
periodic poll cycle.
This only works if traffic is being forwarded by the kernel i.e. if the
entries were HW accelerated via an ASIC we may still miss out on last
minute activity on the mroute in the HW.
Ticket: CM-26871
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
Initially, MLAG Sync is happened at pim_ifchannel, this is mainly to
support even config mismatches(missing configuration of dual active).
But this causes more syncs for each entry.
and also it is not In-line with PIM EVPN. to avoid that moving to
pm_upstream based syncing.
Signed-off-by: Satheesh Kumar K <sathk@cumulusnetworks.com>
(S,G) entries that inherit ipmr-lo into the OIL also inherit
the DF role from the parent (*, G) entry.
This change is done primarily to simplify the sync process and
to prevent the MLAG peers from having to track (S, G) activity etc.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
A local membership is created on the vxlan termination device ipmr-lo. This
is done to -
1. Pull multicast vxlan tunnel traffic to the VTEP for termination by
triggering JoinDesired on the BUM multicast group.
2. Include the OIF in the mroute to signal to the dataplane component
that flow needs to be vxlan terminated.
Earlier we were overloading the PIM_UPSTREAM_FLAG_MASK_SRC_IGMP for
this local membership creation but that is creating confusion both in
the state machine and in the show outputs. To avoid that we use the
more apparent PIM_UPSTREAM_FLAG_MASK_SRC_VXLAN_TERM. With this change -
1. We get LHR functionality for VXLAN_TERM mroutes
2. OIF is populated with PIM_OIF_FLAG_PROTO_PIM only
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
This is not causing functional problems but has become a source
of confusion. DF status is only relevant to multicast tunnel decaps.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
The RPF cost is incremented by 10 if the RPF interface is the peerlink-rif.
This is used to force the MLAG switch with the lowest cost to the RPF
to become the MLAG DF. If a switch has to go via the peerlink-rif to get
to the RP or source it simplly cannot be the designated forwarder.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
DF election is only run for (*,G) entries i.e. election is skipped
for (S,G) entries that are setup as a result of SPT switchover. (S,G)
entries inherit the DF role from the parent (*,G) entry. So the DF is
responsible for terminating all sources associated with a group.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
1. Upstream entries associated with tunnel termination mroutes are
synced to the MLAG peer via the local MLAG daemon.
2. These entries are installed in the peer switch (via an upstream
ref flag).
3. DF (Designated Forwarder) election is run per-upstream entry by both
the MLAG switches -
a. The switch with the lowest RPF cost is the DF winner
b. If both switches have the same RPF cost the MLAG role is
used as a tie breaker with the MLAG primary becoming the DF
winner.
4. The DF winner terminates the multicast traffic by adding the tunnel
termination device to the OIL. The non-DF suppresses the termination
device from the OIL.
Note: Before the PIM-MLAG interface was available hidden config was
used to test the EVPN-PIM functionality with MLAG. I have removed the
code to persist that config to avoid confusion. The hidden commands are
still available.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
Convert the upstream_list and hash to a rb tree, Significant
time was being spent in the listnode_add_sort. This reduces
this time greatly.
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
use_rpt macro depends on JoinDesired macro and is mostly independent of the
actual RPF interface i.e. doesn't change when the RPF interface changes.
There is however one exception to this handling and that is on the
first hop router (DR or non-DR). On the DR the FHR flag is set so the
RPF interface stays irrelevant to use_rpt eval. But on the non-DR the
IIF is the only way to know we are directly connected to the SG i.e.
to know that we must NOT switch the source to RPT.
This commit fixes up the order of use_rpt eval -
1. it is done before mroute programming
2. but after IIF setup, for SRC_NOCACHE and STATIC_IIF upstream entries
Note: drop an unnecessary check to verify that the RPF interface is
pim enabled. This is just to make the code consistent.
Ticket: CM-27446
Signed-off-by: Donald Sharp <sharpd@cumulusnetworks.com>
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
JD macro is defined by the RFC as -
bool JoinDesired(S,G) {
return (immediate_olist(S,G) != NULL
OR (KeepaliveTimer(S,G) is running
AND inherited_olist(S,G) != NULL))
}
However for MSDP synced SA the KAT will not be running so an exception is
needed. Earlier I had done this by relaxing KAT_run requirements entirely
on the RP. However as that prevents the source from being aged out in some
cases I have made the check more narrow i.e. has to an MSDP peer added
entry.
Ticket: CM-24398
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
Added event logs around add/del of upstream entries into the nbr's
jp-agg list. This is to help debug a problem with stale (deleted)
upstream entries being present in the list causing pimd to crash on
the periodic processing.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
Today we are only pruning the SPT when (S,G) upstream entry
switches from Joined toNotJoined. This leaves the source still
pruned along the RPT till the next periodic XG join-prune is sent
to the RPF(RP). Traffic from the source will be blackholed for this
duration. To prevent that we need send a new JP message
to RPF(RP) immediately.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
This was causing pimd to crash later; call-stack -
(gdb) bt
context=<optimized out>) at lib/sigevent.c:254
group=group@entry=0x7ffffa9797e0) at pimd/pim_rp.c:207
grp=grp@entry=0x7ffffa9799fe, sgs=sgs@entry=0x560ac069edb0, size=52)
at pimd/pim_msg.c:200
groups=<optimized out>) at pimd/pim_join.c:562
at pimd/pim_neighbor.c:288
at lib/thread.c:1599
at lib/libfrr.c:1024
envp=<optimized out>) at pimd/pim_main.c:162
(gdb) fr 4
group=group@entry=0x7ffffa9797e0) at pimd/pim_rp.c:207
207 pimd/pim_rp.c: No such file or directory.
(gdb) fr 6
grp=grp@entry=0x7ffffa9799fe, sgs=sgs@entry=0x560ac069edb0, size=52)
at pimd/pim_msg.c:200
200 pimd/pim_msg.c: No such file or directory.
(gdb) p source->up->sg_str
$1 = '\000' <repeats 31 times>, <incomplete sequence \361>
(gdb)
This problem can manifest in the following event sequence -
1. upstream RPF neighbor is resolved
2. upstream RPF neighbor becomes unresolved (but upstream entry
stays on the jp-agg list)
3. upstream entry is removed
on the next old-neighbor jp-agg-list processing the stale entry is
accessed resulting in the crash.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
A dummy pim upstream entry can be in a JOINED state before its RPF nbr is
added. Handle that case by triggering an immediate join.
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
1. KAT should be re-started only if traffic rxed along the SPT i.e.
IIF == RPF_Interface(S).
Only exception to the rule is if you are LHR.
2. KAT should be started on all routers (not just FHR, RP, LHR).
Signed-off-by: Anuradha Karuppiah <anuradhak@cumulusnetworks.com>
