Public bug reported:
System running 5.13.0 kernel, OFED 5.4.3, OVS 2.16 and OVN 21.09.
ConnectX-5 with firmware=16.31.2006
Hardware offload enabled.
Ports bonded (VF LAG enabled).
Security Groups enabled (i.e. CT offload).
When the system is the active gateway chassis for a logical OVN router,
i.e. it is currently responsible for north/south traffic for instances
that may be on other hosts, I see disturbances in the traffic and see
messages like this in the log.
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.761959] mlx5_core 0000:81:00.0:
mlx5dr_actions_build_ste_arr:721:(pid 6628): Failed to handle checksum
recalculation err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.774656] mlx5_core 0000:81:00.0:
dr_rule_create_rule:1287:(pid 6628): Failed creating rule
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.774659] mlx5_core 0000:81:00.0:
mlx5_cmd_dr_create_fte:561:(pid 6628): Failed to create dr rule err(-22)
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.784057] mlx5_core 0000:81:00.0
enp129s0f0: Failed to add post action rule
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.819968] mlx5_core 0000:81:00.0
enp129s0f0: Failed to offload ct flow, err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.829390] mlx5_core 0000:81:00.1:
mlx5dr_actions_build_ste_arr:721:(pid 6628): Failed to handle checksum
recalculation err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.842403] mlx5_core 0000:81:00.1:
dr_rule_create_rule:1287:(pid 6628): Failed creating rule
Nov 3 15:28:35 pc1-rb3-n4 kernel: [ 7065.597150] mlx5_core
0000:81:00.0: mlx5_cmd_check:810:(pid 6654):
DEALLOC_PACKET_REFORMAT_CONTEXT(0x93e) op_mod(0x0) failed, status bad
resource state(0x9), syndrome (0x179e84)
** Affects: linux (Ubuntu)
Importance: Undecided
Status: Incomplete
--
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1949609
Title:
[mlx5][ct-offload] Traffic disruption and errors logged when system is
gateway chassis for OVN logical router
Status in linux package in Ubuntu:
Incomplete
Bug description:
System running 5.13.0 kernel, OFED 5.4.3, OVS 2.16 and OVN 21.09.
ConnectX-5 with firmware=16.31.2006
Hardware offload enabled.
Ports bonded (VF LAG enabled).
Security Groups enabled (i.e. CT offload).
When the system is the active gateway chassis for a logical OVN
router, i.e. it is currently responsible for north/south traffic for
instances that may be on other hosts, I see disturbances in the
traffic and see messages like this in the log.
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.761959] mlx5_core 0000:81:00.0:
mlx5dr_actions_build_ste_arr:721:(pid 6628): Failed to handle checksum
recalculation err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.774656] mlx5_core 0000:81:00.0:
dr_rule_create_rule:1287:(pid 6628): Failed creating rule
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.774659] mlx5_core 0000:81:00.0:
mlx5_cmd_dr_create_fte:561:(pid 6628): Failed to create dr rule err(-22)
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.784057] mlx5_core 0000:81:00.0
enp129s0f0: Failed to add post action rule
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.819968] mlx5_core 0000:81:00.0
enp129s0f0: Failed to offload ct flow, err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.829390] mlx5_core 0000:81:00.1:
mlx5dr_actions_build_ste_arr:721:(pid 6628): Failed to handle checksum
recalculation err -22
Nov 3 13:35:01 pc1-rb3-n4 kernel: [ 250.842403] mlx5_core 0000:81:00.1:
dr_rule_create_rule:1287:(pid 6628): Failed creating rule
Nov 3 15:28:35 pc1-rb3-n4 kernel: [ 7065.597150] mlx5_core
0000:81:00.0: mlx5_cmd_check:810:(pid 6654):
DEALLOC_PACKET_REFORMAT_CONTEXT(0x93e) op_mod(0x0) failed, status bad
resource state(0x9), syndrome (0x179e84)
To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1949609/+subscriptions
--
Mailing list: https://launchpad.net/~kernel-packages
Post to : kernel-packages@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kernel-packages
More help : https://help.launchpad.net/ListHelp
