diff options
| author |   Shawn Ji <xiaji@tethrnet.com> | 2019-12-18 10:10:54 +0800 |
|---|---|---|
| committer | Shawn Ji <xiaji@tethrnet.com> | 2019-12-20 18:25:10 +0800 |
| commit | 623b4f85e6ee4611ae15bb3103fe30725ca977ed (patch) | |
| tree | c0003115d71d300d554d8ba4ac2fee9029d8e731 /src/vnet/geneve/encap.c | |
| parent | 4453c09ac38b8c80248d105e378640230124f8a6 (diff) | |
vxlan: reuse inner packet flow hash for tunnel outer header load balance
Type: fix
Several tunnels encapsulation use udp as outer header and udp src port
is set by inner header flow hash, such as gtpu, geneve, vxlan, vxlan-gbd
Since flow hash of inner header is already been calculated, keeping it
to vnet_buffere[b]->ip.flow_hash should save load-balance node work to
select ECMP uplinks.
Change-Id: I0e4e2b27178f4fcc5785e221d6d1f3e8747d0d59
Signed-off-by: Shawn Ji <xiaji@tethrnet.com>
Diffstat (limited to 'src/vnet/geneve/encap.c')
| -rw-r--r-- | src/vnet/geneve/encap.c | 8 |
1 files changed, 8 insertions, 0 deletions
diff --git a/src/vnet/geneve/encap.c b/src/vnet/geneve/encap.c index 8e59aea3626..af52fd24543 100644 --- a/src/vnet/geneve/encap.c +++ b/src/vnet/geneve/encap.c @@ -120,6 +120,7 @@ geneve_encap_inline (vlib_main_t * vm, flow_hash0 = vnet_l2_compute_flow_hash (b[0]); flow_hash1 = vnet_l2_compute_flow_hash (b[1]); + /* Get next node index and adj index from tunnel next_dpo */ if (sw_if_index0 != vnet_buffer (b[0])->sw_if_index[VLIB_TX]) { @@ -289,6 +290,10 @@ geneve_encap_inline (vlib_main_t * vm, stats_n_packets += 2; stats_n_bytes += len0 + len1; + /* save inner packet flow_hash for load-balance node */ + vnet_buffer (b[0])->ip.flow_hash = flow_hash0; + vnet_buffer (b[1])->ip.flow_hash = flow_hash1; + /* Batch stats increment on the same geneve tunnel so counter is not incremented per packet. Note stats are still incremented for deleted and admin-down tunnel where packets are dropped. It is not worthwhile @@ -467,6 +472,9 @@ geneve_encap_inline (vlib_main_t * vm, stats_n_packets += 1; stats_n_bytes += len0; + /* save inner packet flow_hash for load-balance node */ + vnet_buffer (b[0])->ip.flow_hash = flow_hash0; + /* Batch stats increment on the same geneve tunnel so counter is not incremented per packet. Note stats are still incremented for deleted and admin-down tunnel where packets are dropped. It is not worthwhile |