diff options
| author |   Vladislav Grishenko <themiron@yandex-team.ru> | 2024-01-24 16:17:23 +0500 |
|---|---|---|
| committer |   Neale Ranns <neale@graphiant.com> | 2024-04-09 04:47:02 +0000 |
| commit | 302db471a01b5c21b80e9bc769315cef4da0ed80 (patch) | |
| tree | be284a966c720d6dc601128759e42a62553db2a7 /src/vnet/ip/ip6_forward.c | |
| parent | 2a7bc81ae302b02e8e0d4eec7db82c1ca15b7f7a (diff) | |
mpls: fix default mpls lb hash config
In case of multiple path within tunnel, mpls lookup node
computes lb hash with mpls_compute_flow_hash config value 0,
so only mpls label and l4 ports gets accounted, not 5-tuple.
This leads to flow traffic polarization and disbalance over
mpls paths.
Use mpls hash config from lb instead, usually it'll be
MPLS_FLOw_HASH_DEFAULT with 5-tuple plus flowlabel.
As optimization, fix flow hash reuse from the previous lookup
node if present, like ip_lookup does. Previously mpls lookup
always calcs the hash.
Test lb distribution for both cases.
Also, use the same flow hash hex format in ip4/ip6 and mpls
traces for easier reading, most code changes is due fixstyle
formatting.
Type: fix
Signed-off-by: Vladislav Grishenko <themiron@yandex-team.ru>
Change-Id: Ib89e1ab3edec14269866fe825a3e887d6c817b7c
Diffstat (limited to 'src/vnet/ip/ip6_forward.c')
| -rw-r--r-- | src/vnet/ip/ip6_forward.c | 3 |
1 files changed, 1 insertions, 2 deletions
diff --git a/src/vnet/ip/ip6_forward.c b/src/vnet/ip/ip6_forward.c index 06c473b1495..48fb633fd32 100644 --- a/src/vnet/ip/ip6_forward.c +++ b/src/vnet/ip/ip6_forward.c @@ -948,8 +948,7 @@ format_ip6_forward_next_trace (u8 * s, va_list * args) ip6_forward_next_trace_t *t = va_arg (*args, ip6_forward_next_trace_t *); u32 indent = format_get_indent (s); - s = format (s, "%Ufib:%d adj:%d flow:%d", - format_white_space, indent, + s = format (s, "%Ufib:%d adj:%d flow:0x%08x", format_white_space, indent, t->fib_index, t->adj_index, t->flow_hash); s = format (s, "\n%U%U", format_white_space, indent, |