8:40 a.m.
The next patches in this patch-set are going to introduce several multicast
router tests. Add common topology and common module for these tests.
The module allows to add/remove VIFs and MFCs. In addition, the module can
test multicast forwarding by running traffic and verifying the traffic was
forwarded correctly.
The module follows the kernel API and validates that the added MFC and VIFs
behave as expected. The module tests that:
- The traffic is forwarded according to the route output interfaces.
- If there is a pimreg interface in the route, test that the WHOLEPKT
notifications were sent.
- When packet is sent to a wrong interface, validate that the matching
notification is sent.
Signed-off-by: Yotam Gigi <yotamg(a)mellanox.com>
---
recipes/switchdev/mr_common.py | 205 +++++++++++++++++++++++++++++++
recipes/switchdev/mr_common_topology.xml | 78 ++++++++++++
2 files changed, 283 insertions(+)
create mode 100644 recipes/switchdev/mr_common.py
create mode 100644 recipes/switchdev/mr_common_topology.xml
diff --git a/recipes/switchdev/mr_common.py b/recipes/switchdev/mr_common.py
new file mode 100644
index 0000000..69dcdb6
--- /dev/null
+++ b/recipes/switchdev/mr_common.py
@@ -0,0 +1,205 @@
+from time import sleep
+from lnst.Controller.Task import ctl
+from lnst.Common.Consts import MROUTE
+import copy
+
+# The topology used is:
+#
+# +----------------------------------------------------+
+# | switch |
+# | |
+# |addr: 1.10 2.10 3.10 4.10 5.10 |
+# |port: if1 if2 if3 if4 if5 |
+# | + + + + + |
+# | | | br0 | | | |
+# | | +------+------+ | | | |
+# | | | | | | | |
+# +----------------------------------------------------+
+# | | | | | |
+# | | | | | |
+# +------------------+ +-----------------------------+
+# | | | | | | | | | |
+# | + + | | + + + + |
+# |port if1 if2 | | if1 if2 if3 if4 |
+# |addr: 1.1 2.1 | | 2.2 3.2 4.2 5.2 |
+# | | | |
+# | machine1 | | machine2 |
+# +------------------+ +-----------------------------+
+#
+
+def check_results(sw, tl, ifaces, ratios, expected):
+ results = [r > 0.9 and r < 1.1 for r in ratios]
+ err_indices = [i for i in range(len(results)) if results[i] != expected[i]]
+ if err_indices != []:
+ err_str = "Interfaces that got traffic are %s, while expected are %s" %
\
+ (str(ratios), str(expected))
+ tl.custom(sw, "iperf_mc", err_str)
+
+class MrouteTest:
+ def __init__(self, tl, hosts, ifaces):
+ self.tl = tl
+ self.m1, self.m2, self.sw = hosts
+ m1_if1, m1_if2, m2_if1, m2_if2, m2_if3, m2_if4, sw_if1, \
+ sw_br_m1, sw_br_m2, sw_if3, sw_if4, sw_if5, sw_if2 = ifaces
+
+ self.sw_ports = [sw_if1, sw_if2, sw_if3, sw_if4, sw_if5]
+ self.mach_ports = [m1_if1, m1_if2, m2_if1, m2_if2, m2_if3, m2_if4]
+ self.port2vif = {}
+ self.vif2port = {}
+
+ team_config = '{"runner" : {"name" :
"lacp"}}'
+ self.unrif_lag = self.sw.create_team(slaves = [],
+ config=team_config)
+
+ self.sw_mach_conn = { sw_if1: [m1_if1], sw_if2: [m1_if2, m2_if1],
+ sw_if3: [m2_if2], sw_if4: [m2_if3], sw_if5: [m2_if4]
+ }
+ self.mach_sw_conn = { m1_if1: sw_if1, m1_if2: sw_if2, m2_if1: sw_if2,
+ m2_if2: sw_if3, m2_if3: sw_if4, m2_if4: sw_if5
+ }
+
+ sw_if2.set_br_mcast_snooping(False)
+
+ def init(self):
+ self.sw.mroute_init()
+ self.sw.mroute_pim_init()
+
+ def add_vif(self, port, vif_index):
+ self.port2vif[port] = vif_index
+ self.vif2port[vif_index] = port
+ port.mroute_add_vif(vif_index)
+
+ def del_vif_port(self, port):
+ vif_index = self.port2vif[port]
+ port.mroute_del_vif(vif_index)
+ del self.port2vif[port]
+ del self.vif2port[vif_index]
+
+ def del_vif(self, vif_index):
+ port = self.vif2port[vif_index]
+ port.mroute_del_vif(vif_index)
+ del self.port2vif[port]
+ del self.vif2port[vif_index]
+
+ def fini(self):
+ self.sw.mroute_finish()
+
+ def del_rif(self, port):
+ self.unrif_lag.slave_add(port.get_id())
+
+ def add_rif(self, port):
+ self.unrif_lag.slave_del(port.get_id())
+ port.set_link_up()
+
+ def pimreg_add(self, vif_index):
+ self.sw.mroute_add_vif_reg(vif_index)
+ self.port2vif["pimreg"] = vif_index
+ self.vif2port[vif_index] = "pimreg"
+
+ def expect_mr_notifs(self, notif_type, num_min=1, num_max=1,
+ source_ip = None, source_vif = None, group_ip = None):
+ num_notifs = 0
+ for i in range(num_max):
+ notif = self.tl.expect_mr_notif(self.sw, notif_type,
+ none_ok = True)
+ if notif == None:
+ break
+ num_notifs += 1
+
+ if num_notifs < num_min:
+ self.tl.custom(self.sw, "mroute notif", "did not get
notification")
+
+ def send_mc_traffic(self, group, source_port, num_packets):
+ speed = 1000
+ source_host = source_port.get_host()
+ source_host.sync_resources(modules=["Iperf"])
+ iperf = self.tl._get_iperf_cli_mod_packets(group, num_packets, speed)
+
+ sleep(1)
+ source_port.enable_multicast()
+ tx_stats_before = source_port.link_stats()["tx_packets"]
+ rx_stats_before = [port.link_stats()["rx_mcast"]
+ for port in self.mach_ports]
+ source_host.run(iperf)
+ sleep(1)
+ tx_stats_after = source_port.link_stats()["tx_packets"]
+ rx_stats_after = [port.link_stats()["rx_mcast"]
+ for port in self.mach_ports]
+ source_port.disable_multicast()
+
+ tx_stats = tx_stats_after - tx_stats_before
+ rx_stats = [before - after
+ for before, after in zip(rx_stats_after, rx_stats_before)]
+ return [float(stats)/tx_stats for stats in rx_stats]
+
+ def test_fwd(self, group, source_port, dest_ports, pimreg = False):
+
+ if source_port == "pimreg":
+ return
+ if source_port in dest_ports:
+ dest_ports.remove(source_port)
+
+ # other_ports are all ports but the source
+ source_mach_port = self.sw_mach_conn[source_port][0]
+ dest_mach_ports = [port for port in self.mach_ports \
+ if self.mach_sw_conn[port] in dest_ports]
+ bridged_mach_ports = self.sw_mach_conn[source_port]
+
+ # Ports that are expected to get traffic are:
+ # - Ports that are destination of the route
+ # - Ports that are bridged to the source port
+ # Traffic should never get to the source port, not even if the route
+ # points to it.
+ expected_res = [(port in dest_mach_ports or port in bridged_mach_ports)
+ and port != source_mach_port for port in self.mach_ports]
+ ratio = self.send_mc_traffic(group, source_mach_port, 100)
+ check_results(self.sw, self.tl, self.mach_ports, ratio, expected_res)
+
+ if pimreg:
+ self.expect_mr_notifs(MROUTE.NOTIF_WHOLEPKT, 3, 200)
+
+ def mroute_test(self, mroute):
+ ivif = mroute["ivif"]
+ evifs = mroute["evifs"]
+
+ # if ivif unresolved or pimreg, don't test
+ if ivif not in self.vif2port.keys() or self.vif2port[ivif] ==
"pimreg":
+ return
+
+ source_port = self.vif2port[ivif]
+ dest_ports = [self.vif2port[evif] for evif in evifs
+ if evif in self.vif2port.keys()]
+
+ pimreg = False
+ if "pimreg" in dest_ports:
+ pimreg = True
+ dest_ports.remove("pimreg")
+
+ # run legitimate traffic and check that it is forwarded
+ self.test_fwd(mroute["group"], source_port, dest_ports, pimreg)
+
+ # If the route is (*,G), check RPF failures, as (S,G) route RPF
+ # failures are much more difficult to check
+ if mroute["source"] == "0.0.0.0" and len(dest_ports) != 0:
+ new_source_port = dest_ports[0]
+ new_source_mach_port = self.sw_mach_conn[new_source_port][0]
+ self.send_mc_traffic(mroute["group"], new_source_mach_port, 1)
+ self.tl.expect_mr_notif(self.sw, MROUTE.NOTIF_WRONGVIF)
+
+ def mroute_create(self, source, group, ivif, evifs, test = True):
+ dest_ports = [self.vif2port[evif] for evif in evifs]
+ source_port = self.vif2port[ivif]
+
+ evif_ttls = {evif: 1 for evif in evifs}
+ self.sw.mroute_add_mfc(source, group, ivif, evif_ttls)
+ mroute = {"ivif": ivif, "evifs": copy.deepcopy(evifs),
+ "source": source, "group": group}
+ if test:
+ self.mroute_test(mroute)
+ return mroute
+
+ def mroute_remove(self, mroute, test = True):
+ self.sw.mroute_del_mfc(mroute["source"], mroute["group"],
mroute["ivif"])
+
+ if test:
+ self.test_fwd(mroute["group"],
self.vif2port[mroute["ivif"]], [])
diff --git a/recipes/switchdev/mr_common_topology.xml
b/recipes/switchdev/mr_common_topology.xml
new file mode 100644
index 0000000..610d45d
--- /dev/null
+++ b/recipes/switchdev/mr_common_topology.xml
@@ -0,0 +1,78 @@
+ <network>
+ <host id="machine1">
+ <interfaces>
+ <eth id="if1" label="A">
+ <addresses>
+ <address value="192.168.1.1/24"/>
+ </addresses>
+ </eth>
+ <eth id="if2" label="B">
+ <addresses>
+ <address value="192.168.2.1/24"/>
+ </addresses>
+ </eth>
+ </interfaces>
+ </host>
+ <host id="machine2">
+ <interfaces>
+ <eth id="if1" label="C">
+ <addresses>
+ <address value="192.168.2.2/24"/>
+ </addresses>
+ </eth>
+ <eth id="if2" label="D">
+ <addresses>
+ <address value="192.168.3.2/24"/>
+ </addresses>
+ </eth>
+ <eth id="if3" label="E">
+ <addresses>
+ <address value="192.168.4.2/24"/>
+ </addresses>
+ </eth>
+ <eth id="if4" label="F">
+ <addresses>
+ <address value="192.168.5.2/24"/>
+ </addresses>
+ </eth>
+ </interfaces>
+ </host>
+ <host id="switch">
+ <interfaces>
+ <eth id="if1" label="A">
+ <addresses>
+ <address value="192.168.1.10/24"/>
+ </addresses>
+ </eth>
+ <eth id="if2" label="B"/>
+ <eth id="if3" label="C"/>
+ <bridge id="br0">
+ <addresses>
+ <address value="192.168.2.10/24"/>
+ </addresses>
+ <slaves>
+ <slave id="if2"/>
+ <slave id="if3"/>
+ </slaves>
+ <options>
+ <option name="vlan_filtering" value="1"/>
+ </options>
+ </bridge>
+ <eth id="if4" label="D">
+ <addresses>
+ <address value="192.168.3.10/24"/>
+ </addresses>
+ </eth>
+ <eth id="if5" label="E">
+ <addresses>
+ <address value="192.168.4.10/24"/>
+ </addresses>
+ </eth>
+ <eth id="if6" label="F">
+ <addresses>
+ <address value="192.168.5.10/24"/>
+ </addresses>
+ </eth>
+ </interfaces>
+ </host>
+ </network>
--
2.8.4