Description of problem: Running Uperf UDP workload from pod to pod across worker is dropping packets and seemed to be dropping because of packet fragmentation. UDP packets fragmenting when the payload size is greater than MTU size. If protocol is set to TCP we don't see any packet drop or fragmentation. Version-Release number of selected component (if applicable): OCP: 4.7.0-fc.5 Kernel version: 4.18.0-240.10.1.el8_3.x86_64 OVN version: Version: 0.3.0 Git commit: bc8321aa1afb6b532e73effc0d2bd3212c15a85e Git branch: rhaos-4.7-rhel-8 Go version: go1.15.5 Build date: 2021-01-30 OS/Arch: linux amd64 OVS version: $ ovs-vsctl --version ovs-vsctl (Open vSwitch) 2.13.2 DB Schema 8.2.0 How reproducible: Frequently on any environment Steps to Reproduce: 1. Install OCP cluster 4.7.0-fc.5 2. Run UPerf workload (We are using benchmark-operator) 3. UDP protocol and Payload size more than MTU Actual results: Packets fragmenting and RX drops at the NIC level Expected results: No Fragmentation and no drops as in TCP Additional info: $ ethtool -i ens2f1 driver: i40e version: 2.8.20-k firmware-version: 7.10 0x800051ac 19.0.12 expansion-rom-version: bus-info: 0000:62:00.1 supports-statistics: yes supports-test: yes supports-eeprom-access: yes supports-register-dump: yes supports-priv-flags: yes Cluster network: status: clusterNetwork: - cidr: 10.128.0.0/14 hostPrefix: 23 clusterNetworkMTU: 8900 networkType: OVNKubernetes serviceNetwork: - 172.30.0.0/16 $ ifconfig ens2f1 ens2f1: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 9000 ether 40:a6:b7:00:47:81 txqueuelen 1000 (Ethernet) RX packets 752005677 bytes 5000469924153 (4.5 TiB) RX errors 0 dropped 83514476 overruns 0 frame 0 TX packets 156543237 bytes 621736118212 (579.0 GiB) TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0 $ ethtool -k ens2f1 Features for ens2f1: rx-checksumming: on tx-checksumming: on tx-checksum-ipv4: on tx-checksum-ip-generic: off [fixed] tx-checksum-ipv6: on tx-checksum-fcoe-crc: off [fixed] tx-checksum-sctp: on scatter-gather: on tx-scatter-gather: on tx-scatter-gather-fraglist: off [fixed] tcp-segmentation-offload: on tx-tcp-segmentation: on tx-tcp-ecn-segmentation: on tx-tcp-mangleid-segmentation: off tx-tcp6-segmentation: on generic-segmentation-offload: on generic-receive-offload: on large-receive-offload: off [fixed] rx-vlan-offload: on tx-vlan-offload: on ntuple-filters: on receive-hashing: on highdma: on rx-vlan-filter: on [fixed] vlan-challenged: off [fixed] tx-lockless: off [fixed] netns-local: off [fixed] tx-gso-robust: off [fixed] tx-fcoe-segmentation: off [fixed] tx-gre-segmentation: on tx-gre-csum-segmentation: on tx-ipxip4-segmentation: on tx-ipxip6-segmentation: on tx-udp_tnl-segmentation: on tx-udp_tnl-csum-segmentation: on tx-gso-partial: on tx-sctp-segmentation: off [fixed] tx-esp-segmentation: off [fixed] tx-udp-segmentation: on tls-hw-rx-offload: off [fixed] fcoe-mtu: off [fixed] tx-nocache-copy: off loopback: off [fixed] rx-fcs: off [fixed] rx-all: off [fixed] tx-vlan-stag-hw-insert: off [fixed] rx-vlan-stag-hw-parse: off [fixed] rx-vlan-stag-filter: off [fixed] l2-fwd-offload: off hw-tc-offload: on esp-hw-offload: off [fixed] esp-tx-csum-hw-offload: off [fixed] rx-udp_tunnel-port-offload: on tls-hw-tx-offload: off [fixed] rx-gro-hw: off [fixed] tls-hw-record: off [fixed]
It seems like this is a dupe of https://bugzilla.redhat.com/show_bug.cgi?id=1927047 I am not sure though so I will hold off doing so until I get trozet's input.
There's not a lot of info in this bugzilla, but I'm fairly certain it is the same as 1927047. *** This bug has been marked as a duplicate of bug 1927047 ***