srv6: Introduce IPv6 address in segment-list used by TE Policy #15057

dmytroshytyi-6WIND · 2023-12-21T17:07:27Z

These series of patches introduce new feature: user may configure ipv6-address in segment list.

Format:

index INDEX ipv6-address X:X::X:X

Example:

segment-routing
 traffic-eng
  segment-list SL1
   index 0 ipv6-address 2001:db8::1

Signed-off-by: Dmytro Shytyi [email protected]

dmytroshytyi-6WIND · 2023-12-26T12:05:45Z

Fix warnings by checkpatch.

dmytroshytyi-6WIND · 2023-12-26T15:06:29Z

2 small fixes for isis_sr_te uploaded in the recent push.

tests/topotests/isis_srv6_te_topo1/dst/zebra.conf

ton31337 · 2023-12-22T08:33:39Z

tests/topotests/isis_srv6_te_topo1/rt1/isisd.conf

@@ -0,0 +1,35 @@
+password 1


Is this needed?

tests/topotests/isis_srv6_te_topo1/rt1/step1/show_ip_route.ref

tests/topotests/isis_srv6_te_topo1/test_isis_srv6_te_topo1.py

zebra/zebra_srv6.c

tests/topotests/isis_srv6_te_topo1/rt1/zebra.conf

tests/topotests/isis_srv6_te_topo1/rt2/zebra.conf

tests/topotests/isis_srv6_te_topo1/rt3/zebra.conf

tests/topotests/isis_srv6_te_topo1/rt4/zebra.conf

ton31337 · 2024-01-06T18:37:55Z

tests/topotests/isis_srv6_te_topo1/test_isis_srv6_te_topo1.py

+    create_interface_in_kernel,
+)
+
+pytestmark = [pytest.mark.isisd, pytest.mark.sharpd]


Shouldn't bgpd, pathd be included also?

this is not fixed.

really fixed by also adding the

pytest.mark.pathd

riw777

some questions; waiting on @ton31337 's comments as well

riw777 · 2024-01-16T14:43:53Z

lib/zclient.c

+
+	stream_putw(s, zt->srv6_segs.num_segs);
+
+	stream_put(s, &zt->srv6_segs.segs[0],


yes ... this should check for an empty seg before putting, I think

lib/zclient.c

tests/topotests/isis_srv6_te_topo1/rt1/step1/show_ipv6_route.ref

lib/srte.h

donaldsharp · 2024-01-16T17:53:56Z

lib/srte.h

+	ZEBRA_SR_SRV6_SRTE = 2, /* SRv6 SID List*/
+};
+
+static inline enum lsp_types_t lsp_type_from_sr_type(int sr_type)


Shouldn't this pass in a enum sr_types.. instead of an int? Let's let the compiler help us please

Please remove the default case here.

pathd/path_zebra.c

donaldsharp · 2024-01-16T18:27:31Z

zebra/rt_netlink.c

@@ -1550,7 +1550,7 @@ static ssize_t fill_seg6ipt_encap(char *buffer, size_t buflen,
 	srh->first_segment = segs->num_segs - 1;

 	for (i = 0; i < segs->num_segs; i++) {
-		memcpy(&srh->segments[i], &segs->seg[i],
+		memcpy(&srh->segments[segs->num_segs - i - 1], &segs->seg[i],


This is only 1/2 of the equation, we need to fix the decode side as well. Add the ipv6 seg6 route from the linux cli and the segments are backwards in comparison.

Thanks, fixed,

donaldsharp

Some work needs to be done to clean this up

pathd/path_zebra.c

donaldsharp · 2024-01-16T20:19:27Z

zebra/zebra_srte.c

+
+		for (rn = route_top(table); rn; rn = route_next(rn)) {
+			RNODE_FOREACH_RE_SAFE (rn, re, next) {
+				if (re->nhe->nhg.nexthop->srte_color ==


invert this logic so the indent is not so deep this is hard to read

donaldsharp · 2024-01-16T20:20:17Z

zebra/zebra_srte.c

@@ -133,6 +136,47 @@ static int zebra_sr_policy_notify_update_client(struct zebra_sr_policy *policy,
 	}
 	stream_putl(s, policy->color);

+	if (policy->segment_list.srv6_segs.num_segs > 0) {
+		/* Lookup table.  */
+		table = zebra_vrf_table(AFI_IP6, SAFI_UNICAST, VRF_DEFAULT);


What is supposed to happen when we have > 1 million routes? What is a reasonable number of items that are going to be sent. Are we going to run out of packet space? Are we going to just cpuhog like you would not believe?

tests/topotests/isis_srv6_te_topo1/test_isis_srv6_te_topo1.py

dmytroshytyi-6WIND · 2024-01-18T12:48:12Z

Thank you very much for your reviews, @ton31337, @riw777, @donaldsharp.

I was on PTO, thus the response delay. I expect to look at this PR in coming time.

@cscarpitta, WDYT about the cli format for ipv6 address? Does it look good?

donaldsharp · 2024-08-15T12:51:10Z

lib/srte.h

+		return ZEBRA_LSP_SRTE;
+	case ZEBRA_SR_LSP_NONE:
+	case ZEBRA_SR_SRV6_SRTE:
+	default:


remove the default this is an enum and all the values should be accounted for in the switch statement. Leaving a default allows future changes to the enum to be masked here and the developer will miss it.

donaldsharp · 2024-08-15T12:57:54Z

pathd/pathd.c

@@ -621,7 +622,6 @@ void srte_policy_apply_changes(struct srte_policy *policy)
 			SET_FLAG(new_best_candidate->flags, F_CANDIDATE_BEST);
 			SET_FLAG(new_best_candidate->flags,
 				 F_CANDIDATE_MODIFIED);
-


why the whitespace change?

donaldsharp · 2024-08-15T13:18:41Z

When I run the new test under any type of load it fails badly:

================================================================== short test summary info ==================================================================
FAILED ospf6_ecmp_inter_area/test_h.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED ospf6_ecmp_inter_area/test_e.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED isis_srv6_te_topo1/test_zy.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zs.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zd.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_ze.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zz.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zr.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zi.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zc.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zh.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zv.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_isis_srv6_te_topo1.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zt.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zn.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zm.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zk.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zo.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zw.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zj.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zb.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zu.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_za.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zp.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zf.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zl.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zq.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED ospf6_ecmp_inter_area/test_b.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED isis_srv6_te_topo1/test_zx.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
================================================= 29 failed, 1732 passed, 279 skipped in 819.59s (0:13:39) =============================```

Please spend some time running your new test under different loads and identifying and fixing the problems found.

donaldsharp · 2024-08-15T13:31:16Z

as a note when I run the test by itself or concurrently with itself it works fine. It's only when scale/load is placed on the system that causes it to fail badly. This means, to me, that the test is starting to look for data before some other event has occurred and we get this output. See the hundreds of fixes in the topotest directory as we fix this sort of problem in the test for guidance on how to approach the problem. #16586, #16583, #16523, #16510 and #16485 as recent examples of this sort of problem.

dmytroshytyi-6WIND · 2024-08-19T13:07:37Z

When I run the new test under any type of load it fails badly:

================================================================== short test summary info ==================================================================
FAILED ospf6_ecmp_inter_area/test_h.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED ospf6_ecmp_inter_area/test_e.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED isis_srv6_te_topo1/test_zy.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zs.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zd.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_ze.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zz.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zr.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zi.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zc.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zh.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zv.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_isis_srv6_te_topo1.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zt.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zn.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zm.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zk.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zo.py::test_srv6_te_policy_removed - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zw.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zj.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zb.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zu.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_za.py::test_rib_ipv6_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zp.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zf.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zl.py::test_srv6_te_policy_activated - AssertionError: "rt1" JSON output mismatches the expected result
FAILED isis_srv6_te_topo1/test_zq.py::test_srv6_locator_step1 - AssertionError: "rt1" JSON output mismatches the expected result
FAILED ospf6_ecmp_inter_area/test_b.py::test_ecmp_inter_area - AssertionError: 'r1' wrong number of route nexthops
FAILED isis_srv6_te_topo1/test_zx.py::test_srv6_te_policy_additional_route - AssertionError: "rt1" JSON output mismatches the expected result
================================================= 29 failed, 1732 passed, 279 skipped in 819.59s (0:13:39) =============================```

Please spend some time running your new test under different loads and identifying and fixing the problems found.

Hi @donaldsharp , thanks for this comment. Are you running the topotests on the upstream CI hardware?
Reproduction attempt on my local machine of the concurrent test (including parallel load and many tests) run, doesn't face this failures.

============================= test session starts ==============================
platform linux -- Python 3.10.12, pytest-8.2.2, pluggy-1.5.0 -- /usr/bin/python3
cachedir: .pytest_cache
rootdir: /home/frr/frr/tests/topotests
configfile: pytest.ini
plugins: xdist-3.6.1
created: 12/12 workers
12 workers [7206 items]

scheduling tests via LoadFileScheduling

=============================== warnings summary ===============================
..
[gw8] PASSED isis_srv6_te_topo1/test_isis_srv6_te_topo1af.py::test_srv6_te_policy_removed 
isis_srv6_te_topo1/test_isis_srv6_te_topo1af.py::test_memory_leak 
[gw1] SKIPPED isis_srv6_te_topo1/test_isis_srv6_te_topo1ag.py::test_memory_leak 
[gw5] PASSED isis_srv6_te_topo1/test_isis_srv6_te_topo1ai.py::test_srv6_te_policy_removed 
isis_srv6_te_topo1/test_isis_srv6_te_topo1ai.py::test_memory_leak 
[gw7] SKIPPED isis_srv6_te_topo1/test_isis_srv6_te_topo1ah.py::test_memory_leak 
[gw8] SKIPPED isis_srv6_te_topo1/test_isis_srv6_te_topo1af.py::test_memory_leak 
[gw5] SKIPPED isis_srv6_te_topo1/test_isis_srv6_te_topo1ai.py::test_memory_leak 
[gw9] PASSED bgp_local_asn_dot/test_bgp_local_asn_dot_topo1.py::test_verify_bgp_local_as_allow_as_in_iBGP_p0 
bgp_local_asn_dot/test_bgp_local_asn_dot_topo1.py::test_verify_bgp_local_as_in_EBGP_port_reset_p0 
[gw6] PASSED multicast_pim_sm_topo1/test_multicast_pim_sm_topo1.py::test_modify_igmp_query_timer_p0 
multicast_pim_sm_topo1/test_multicast_pim_sm_topo1.py::test_modify_igmp_max_query_response_timer_p0 
isis_srv6_te_topo1/test_isis_srv6_te_topo1al.py::test_isis_adjacencies_step1 
isis_srv6_te_topo1/test_isis_srv6_te_topo1am.py::test_isis_adjacencies_step1 
isis_srv6_te_topo1/test_isis_srv6_te_topo1an.py::test_isis_adjacencies_step1 
isis_srv6_te_topo1/test_isis_srv6_te_topo1ao.py::test_isis_adjacencies_step1 
[gw0] PASSED multicast_pim_uplink_topo1/test_multicast_pim_uplink_topo1.py::test_mroutes_updated_correctly_after_source_interface_shut_noshut_p1 
multicast_pim_uplink_topo1/test_multicast_pim_uplink_topo1.py::test_mroutes_updated_correctly_after_receiver_interface_shut_noshut_p1 
[gw3] PASSED ospfv3_basic_functionality/test_ospfv3_authentication.py::test_ospf6_auth_trailer_tc8_keychain_sha256_missmatch 
ospfv3_basic_functionality/test_ospfv3_authentication.py::test_ospf6_auth_trailer_tc9_keychain_not_configured 
[gw4] PASSED isis_srv6_te_topo1/test_isis_srv6_te_topo1aj.py::test_isis_adjacencies_step1 
isis_srv6_te_topo1/test_isis_srv6_te_topo1aj.py::test_configure_srv6_locators 
[gw4] PASSED isis_srv6_te_topo1/test_isis_srv6_te_topo1aj.py::test_configure_srv6_locators 
isis_srv6_te_topo1/test_isis_srv6_te_topo1aj.py::test_rib_ipv6_step1 
[test.txt](https://github.com/user-attachments/files/16660557/test.txt)

..
========= 6399 passed, 808 skipped, 10 warnings in 4328.23s (1:12:08) ==========

Best Dmytro

dmytroshytyi-6WIND · 2024-08-19T17:15:00Z

as a note when I run the test by itself or concurrently with itself it works fine. It's only when scale/load is placed on the system that causes it to fail badly. This means, to me, that the test is starting to look for data before some other event has occurred and we get this output. See the hundreds of fixes in the topotest directory as we fix this sort of problem in the test for guidance on how to approach the problem. #16586, #16583, #16523, #16510 and #16485 as recent examples of this sort of problem.

Hello @donaldsharp , I have checked all PRs you suggested:

#16586 :

uses topotest.run_and_expect(..) to ensure that Connected Routes are actually installed
timeout 30

#16586 :
This test uses the same tool but with bigger timeout, so let's stick to higher timeouts then.

topotest.run_and_expect(..) +output = net["r1"]
timeout 60

#16523 :
This test introduce different way to use show. Two previous (majority stick to run_and_expect (so run_and_expect is selected))

@retry(retry_timeout=10)
def check_client_connect(r1):
    out = r1.vtysh_cmd("show mgmt backend-adapter all")
    assert "mgmtd-testc" in out

#16510:
Apply timers in the PR's topotest

Timeout 60.
Sets timers to BGP in configuration

#16485:

Addresses flapping routes, not sure if the current checks are a subject of being changed in this direction. They conform with firs two topotests.

Summary:
-+-+-+-+-

Timeout was set to 120 for all tests in this PR.
Timers were minimized in BGP and ISIS.
The difference I see between this PR and listed in quoted message - the way to configure the vtysh:
this PR is not using:

 output = net["r1"].cmd(
            'vtysh -c "show ip route {} nexthop-group"'.format(route_str)
        )

but instead uses:

    tgen.gears["rt1"].vtysh_cmd(
        "configure \n \
        no ipv6 route 2001:db8:10::/64 fc00:0:6:: color 1 \n \
        ipv6 route fc00:0:6b::/48 fc00:0:6:: color 1 \n \
        exit \
        !"
    )

Has someone encountered similar behavior when configuring in this way? I would do and verify the change, but have a setup that doesn't reproduce mentioned issue (regarding the failed tests under the load).

dmytroshytyi-6WIND · 2024-08-20T15:07:29Z

Failures are unrelated:

IPv4:

RFC Failure: MUST

Peer 192.168.0.101 did not forward MPLS packet with label 17
============== Reference: ============================
RFC 3036, s2.7 p23 LDP Identifiers and Next Hop Addresses
============== Test Summary: =========================
Similarly, when the LSR learns a label for a prefix from an LDP peer,
it must be able to determine whether that peer is currently a next hop
(3 more lines...)

ospf

test_ospfv3_single_area test_ospfv3_hello_tc10_p0 | 1 min
AssertionError: Testcase test_ospfv3_hello_tc10_p0 [22]: Failed     Error: [DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1 assert '[DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1' is True E   AssertionError: Testcase test_ospfv3_hello_tc10_p0 [22]: Failed         Error: [DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1     assert '[DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1' is True
1 min
AssertionError: Testcase test_ospfv3_hello_tc10_p0 [22]: Failed 
   Error: [DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1
assert '[DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1' is True
E   AssertionError: Testcase test_ospfv3_hello_tc10_p0 [22]: Failed 
       Error: [DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1
    assert '[DUT: r0] missing OSPF neighbor r1 with router-id 100.1.1.1' is True

riw777 · 2024-10-08T12:25:37Z

I think we're still waiting on @donaldsharp to clear his review on this one ... and the CI changes ... has this been rebased in a while? It might help to rebase it

Add SRv6 segments to zapi_srte_stunnel structure. Signed-off-by: Dmytro Shytyi <[email protected]>

Add SRv6 SID value (srv6-sid-value) to the YANG model. Implement nb handlers for this new value. Signed-off-by: Dmytro Shytyi <[email protected]>

Add structute sr_types_t, that contains the Segment Routing types. Fill the zapi_sr_policy's segment_list with values such as: - value - local_label - num_segs - srv6_segs Signed-off-by: Dmytro Shytyi <[email protected]>

SR-TE tunnel must contain at least one label or SRv6 SID. Check this condition. Signed-off-by: Dmytro Shytyi <[email protected]>

When the segment list is not empty add SRv6 segments to a nexthop. Signed-off-by: Dmytro Shytyi <[email protected]>

Encode and Decode SRv6-TE. It includes: - number of segments - list of segments Signed-off-by: Dmytro Shytyi <[email protected]>

When configuring an SRv6 SID list via iproute2, the segment list that appears in vtysh, learned from the kernel, is inversed. Fix the order of SRv6 SIDs in segment list, learned from the kernel. Fixes: f20cf14 ("bgpd,lib,sharpd,zebra: srv6 introduce multiple segs/SIDs in nexthop") Signed-off-by: Dmytro Shytyi <[email protected]>

Pathd needs to track the reachability of the first SID of any SRv6 segment-list. Use the nexthop tracking facility and add the pathd daemon as a client to nht service from zebra. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

The Pathd daemon is aware of the SID reachability, but still once the sr policy installed in zebra, there is need to re-resolve the SID reachability. Extend the Pathd NHT service, by extracting the resolved next-hop. This information will be conveyed in the SR_POLICY_SET message, and will be very helpful to simplify code in zebra. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

When SRv6 list of segments was changed, change SRv6 policy. Signed-off-by: Dmytro Shytyi <[email protected]>

…nt_list When notifying an update for a colored nexthop, ZEBRA sends the default PATHD distance value and the 0 metric value. Those two values break the original values that originated from an IGP route, for instance. When transmitting this information to BGP for instance, the distance and values are very important when computing BGP paths. Change this by keeping the original distance and metric values. When SRv6 segment list size is bigger than 0, send segments and other elements via zserv_send_message(). The ZEBRA SRTE code can not rely on the routing table. Use instead the resolved nexthop information gathered by th SRv6 SRTE information. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

Consider an srv6 policy active, when sr policy set is received. The sr_policy_validate() function is however called, the default state value is down, and immediately sent to UP. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

Add pathd srv6 segment list support. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

SRv6 Policy is an ordered list of segments that represent a source-routed policy. Packet flows are steered into an SRv6 Policy. Signed-off-by: Dmytro Shytyi <dmytro@[email protected]> Signed-off-by: Philippe Guibert <[email protected]>

riw777 · 2024-11-12T13:15:18Z

CI failures look related ...

dmytroshytyi-6WIND · 2024-11-12T13:18:14Z

Yes, it is related , after rabase.
I will take a look.

Dmytro

frrbot bot added bugfix documentation libfrr yang zebra labels Dec 21, 2023

github-actions bot added master size/XXL labels Dec 21, 2023

dmytroshytyi-6WIND changed the title ~~Introduce IPv6 address in segment-list used by SRv6 TE Policy~~ srv6: Introduce IPv6 address in segment-list used by TE Policy Dec 22, 2023

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch 2 times, most recently from 3f63bf8 to fccc0c2 Compare December 26, 2023 12:04

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch 2 times, most recently from 4d285ff to 331f8d4 Compare December 26, 2023 15:04

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch from 331f8d4 to 57fdef4 Compare December 26, 2023 15:21

riw777 requested review from ton31337 and riw777 and removed request for ton31337 January 2, 2024 15:41

ton31337 reviewed Jan 6, 2024

View reviewed changes

riw777 reviewed Jan 16, 2024

View reviewed changes

donaldsharp reviewed Jan 16, 2024

View reviewed changes

lib/srte.h Outdated Show resolved Hide resolved

donaldsharp reviewed Jan 16, 2024

View reviewed changes

pathd/path_zebra.c Outdated Show resolved Hide resolved

donaldsharp reviewed Jan 16, 2024

View reviewed changes

pathd/path_zebra.c Outdated Show resolved Hide resolved

donaldsharp reviewed Jan 16, 2024

View reviewed changes

donaldsharp requested changes Jan 16, 2024

View reviewed changes

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch 3 times, most recently from c0b8fb3 to 8f81927 Compare February 15, 2024 13:47

donaldsharp reviewed Aug 15, 2024

View reviewed changes

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch 2 times, most recently from b0fa3ff to 19600f8 Compare August 20, 2024 11:49

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch 2 times, most recently from 9212f6b to 3a41cc3 Compare November 7, 2024 15:19

dmytroshytyi-6WIND and others added 14 commits November 7, 2024 18:51

zclient: add srv6 segs to zapi_srte_tunnel

06adee6

Add SRv6 segments to zapi_srte_stunnel structure. Signed-off-by: Dmytro Shytyi <[email protected]>

yang,pathd: add srv6-sid-value and nb handlers

61960ea

Add SRv6 SID value (srv6-sid-value) to the YANG model. Implement nb handlers for this new value. Signed-off-by: Dmytro Shytyi <[email protected]>

lib,pathd,zebra: fill segment_list with sr types

934efc2

Add structute sr_types_t, that contains the Segment Routing types. Fill the zapi_sr_policy's segment_list with values such as: - value - local_label - num_segs - srv6_segs Signed-off-by: Dmytro Shytyi <[email protected]>

zebra: zapi, at least one label or srv6 sid check

439d3ad

SR-TE tunnel must contain at least one label or SRv6 SID. Check this condition. Signed-off-by: Dmytro Shytyi <[email protected]>

zebra: fill nexthop with srv6 segs same as for labels

5552227

When the segment list is not empty add SRv6 segments to a nexthop. Signed-off-by: Dmytro Shytyi <[email protected]>

zebra: encode, decode srv6 segs

be50fcd

Encode and Decode SRv6-TE. It includes: - number of segments - list of segments Signed-off-by: Dmytro Shytyi <[email protected]>

zebra: change srv6 policy based on segs change.

d39dcb5

When SRv6 list of segments was changed, change SRv6 policy. Signed-off-by: Dmytro Shytyi <[email protected]>

doc: add pathd srv6 segment-list support

443d036

Add pathd srv6 segment list support. Signed-off-by: Philippe Guibert <[email protected]> Signed-off-by: Dmytro Shytyi <[email protected]>

topotest: isis_srv6_te_topo1

5db1c40

SRv6 Policy is an ordered list of segments that represent a source-routed policy. Packet flows are steered into an SRv6 Policy. Signed-off-by: Dmytro Shytyi <dmytro@[email protected]> Signed-off-by: Philippe Guibert <[email protected]>

dmytroshytyi-6WIND force-pushed the srv6_te_policy branch from 3a41cc3 to 5db1c40 Compare November 7, 2024 17:53

pguibert6WIND mentioned this pull request Dec 10, 2024

Support bgp l3vpn over srv6 te policy #17609

Open

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

srv6: Introduce IPv6 address in segment-list used by TE Policy #15057

srv6: Introduce IPv6 address in segment-list used by TE Policy #15057

dmytroshytyi-6WIND commented Dec 21, 2023 •

edited

Loading

dmytroshytyi-6WIND commented Dec 26, 2023

dmytroshytyi-6WIND commented Dec 26, 2023

ton31337 Dec 22, 2023

dmytroshytyi-6WIND Feb 15, 2024

ton31337 Jan 6, 2024

dmytroshytyi-6WIND Feb 15, 2024

donaldsharp Aug 15, 2024

dmytroshytyi-6WIND Aug 20, 2024 •

edited

Loading

riw777 left a comment

riw777 Jan 16, 2024

donaldsharp Jan 16, 2024

dmytroshytyi-6WIND Feb 15, 2024

donaldsharp Aug 15, 2024

dmytroshytyi-6WIND Aug 20, 2024

donaldsharp Jan 16, 2024

dmytroshytyi-6WIND Feb 15, 2024

donaldsharp left a comment

donaldsharp Jan 16, 2024

dmytroshytyi-6WIND Feb 15, 2024

donaldsharp Jan 16, 2024

dmytroshytyi-6WIND commented Jan 18, 2024

donaldsharp Aug 15, 2024

dmytroshytyi-6WIND Aug 20, 2024

donaldsharp Aug 15, 2024

dmytroshytyi-6WIND Aug 20, 2024

donaldsharp commented Aug 15, 2024

donaldsharp commented Aug 15, 2024

dmytroshytyi-6WIND commented Aug 19, 2024

dmytroshytyi-6WIND commented Aug 19, 2024 •

edited

Loading

dmytroshytyi-6WIND commented Aug 20, 2024

riw777 commented Oct 8, 2024

riw777 commented Nov 12, 2024

dmytroshytyi-6WIND commented Nov 12, 2024


		stream_putw(s, zt->srv6_segs.num_segs);

		stream_put(s, &zt->srv6_segs.segs[0],

srv6: Introduce IPv6 address in segment-list used by TE Policy #15057

Are you sure you want to change the base?

srv6: Introduce IPv6 address in segment-list used by TE Policy #15057

Conversation

dmytroshytyi-6WIND commented Dec 21, 2023 • edited Loading

dmytroshytyi-6WIND commented Dec 26, 2023

dmytroshytyi-6WIND commented Dec 26, 2023

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dmytroshytyi-6WIND Aug 20, 2024 • edited Loading

Choose a reason for hiding this comment

riw777 left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

donaldsharp left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

dmytroshytyi-6WIND commented Jan 18, 2024

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

donaldsharp commented Aug 15, 2024

donaldsharp commented Aug 15, 2024

dmytroshytyi-6WIND commented Aug 19, 2024

dmytroshytyi-6WIND commented Aug 19, 2024 • edited Loading

dmytroshytyi-6WIND commented Aug 20, 2024

riw777 commented Oct 8, 2024

riw777 commented Nov 12, 2024

dmytroshytyi-6WIND commented Nov 12, 2024

dmytroshytyi-6WIND commented Dec 21, 2023 •

edited

Loading

dmytroshytyi-6WIND Aug 20, 2024 •

edited

Loading

dmytroshytyi-6WIND commented Aug 19, 2024 •

edited

Loading