wlb: T9472: fix the inverted ttl health test

The ttl test considered the interface healthy whenever ping failed. A dead
uplink, where nothing answers, therefore stayed ACTIVE, while a target
answering the ping marked the interface FAILED.

The test now passes if the hop where the TTL runs out reports it, or if the
target is closer and answers itself. If nothing answers within resp-time,
the test fails.

1 files changed, 7 insertions, 4 deletions

diff --git a/python/vyos/wanloadbalance.py b/python/vyos/wanloadbalance.py
index a8cd87350..faea469c2 100644
--- a/python/vyos/wanloadbalance.py
+++ b/python/vyos/wanloadbalance.py
@@ -17,6 +17,7 @@
 import os
 
 from vyos.defaults import directories
+from vyos.utils.process import rc_cmd
 from vyos.utils.process import run
 
 dhclient_lease = 'dhclient_{0}.lease'
@@ -172,10 +173,12 @@ def health_ping_host(host, ifname, count=1, wait_time=0):
     rc = run(cmd_str)
     return rc == 0
 
-def health_ping_host_ttl(host, ifname, count=1, ttl_limit=0):
-    cmd_str = f'ping -c {count} -t {ttl_limit} -I {ifname} {host}'
-    rc = run(cmd_str)
-    return rc != 0
+def health_ping_host_ttl(host, ifname, count=1, ttl_limit=0, wait_time=0):
+    # The path is up to ttl_limit hops if the hop where the TTL runs out
+    # reports it, or if the target itself is closer and answers
+    cmd_str = f'ping -c {count} -W {wait_time} -t {ttl_limit} -I {ifname} {host}'
+    rc, out = rc_cmd(cmd_str)
+    return rc == 0 or 'Time to live exceeded' in out
 
 def parse_dhcp_nexthop(ifname):
     lease_file = os.path.join(directories['isc_dhclient_dir'], dhclient_lease.format(ifname))