Re: [ovs-discuss] OFPFC_DELETE and OFPFC_DELETE_STRICT commands don't delete openflow rules
On Wed, Jul 14, 2021 at 06:34:46PM +0200, Alejandro Llorens Carrodeguas wrote: > Were having trouble deleting OpenFlow rules using the OFPFC_DELETE or > OFPFC_DELETE_STRICT commands that send a Ryu controller to the OVS switch. Thanks for the report. So far, you've just told us that these commands don't work for your case. I can assure you that they do work in the cases we know about. So, we will need an example or a way to reproduce the problem. The easiest you make it for us, the easier it will be to understand the problem you're seeing and either explain why your expectations are wrong or to fix the problem. ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
Re: [ovs-discuss] How to use masked_set_action?
On Wed, Jul 14, 2021 at 08:51:44AM +0800, taoyunupt wrote: > Hi, > I found some info from vswitch.xml，it says "Masked data can improve > performance by allowing megaflows to match on fewer fields." > I do not know What kinds of flow atcion will use 'masked_set_action'. > Can you give some demo flows and situation？ Anything that sets some fields in a header but not others. For example, setting the IP destination address (but not other IP fields) will use it. ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
[ovs-discuss] [ovn/ovs] OVS hardware offloading
Hi, I am using OVN with openstack neutron on ubuntu 20.04. I want to enable hardware offloading on ovs. I have used below command to to enable offload. ovs-vsctl set Open_vSwitch . other_config:hw-offload=true By reviewing below URL. https://www.openvswitch.org/support/ovscon2019/day2/0951-hw_offload_ovs_con_19-Oz-Mellanox.pdf I just want to enable kernel offload using tc. I am using Broadcom BCM57504 NetXtreme-E and below firmware and driver details. driver: bnxt_en version: 1.10.0 firmware-version: 218.0.169.2/pkg 21.80.16.95 expansion-rom-version: bus-info: :63:00.0 supports-statistics: yes supports-test: yes supports-eeprom-access: yes supports-register-dump: no supports-priv-flags: no I have checked with ethtool -k with on NIC, hw-tc-offload: on is on. When I have enabled offloading. I have started seeing below errors. [ 1197.506839] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x1282 error 0x2 [ 1197.507086] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe730da8000 error=-22 [ 1197.516042] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x127f error 0x2 [ 1197.525717] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe730da8000 error=-22 [ 1227.522630] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x12f7 error 0x2 [ 1227.531524] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1227.540403] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x12f4 error 0x2 [ 1227.548676] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1257.563517] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x136c error 0x2 [ 1257.572898] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1257.583457] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x1369 error 0x2 [ 1257.591647] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 Please advise how to fix it. -- Regards, Ammad ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
[ovs-discuss] [OVN] OVN NB "NAT" to accept CIDRs in logical_ip
Hello all: First of all, a bit of context. This question/request is related to [1]. The problem we have in OpenStack is that we are trying to create a VIP but not with a /32 IP address but with a CIDR. Since [2][3], is it possible to define VIPs in Neutron using OVN by creating an unassigned port. This port will provide the VIP. When this VIP is set to a bound port as "allowed_address_pairs", we set the LSP port type to "virtual". In Neutron we can assign a CIDR to a port as an "allowed_address_pairs". For example: $ openstack port show port10 | grep allowed_address_pairs allowed_address_pairs | ip_address='1.2.3.0/24', mac_address='fa:16:3e:7d:ca:e4' The problem is that OVN, to set the virtual LSP chassis and thus create the OpenFlow rules in the selected OVS, is expecting an ARP from the exact IP address that matches with the VIP IP address. Any other IP address in the CIDR '1.2.3.0/24' (following the example provided), won't trigger the rule creation. Is there any mechanism to provide support for a virtual CIDR instead of a VIP? If not, are there any plans for this? Thank you in advance. Rodolfo Alonso (ralon...@redhat.com). [1]https://bugzilla.redhat.com/show_bug.cgi?id=1970907 [2] https://github.com/ovn-org/ovn/commit/054f4c85c413e20d893e10ba053ec52ac15db49c [3]https://review.opendev.org/c/openstack/networking-ovn/+/676223 ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
[ovs-discuss] OFPFC_DELETE and OFPFC_DELETE_STRICT commands don't delete openflow rules
Hello, Were having trouble deleting OpenFlow rules using the OFPFC_DELETE or OFPFC_DELETE_STRICT commands that send a Ryu controller to the OVS switch. What you did that make the problem appear: The Ryu controller sends OFPFC_DELETE or OFPFC_DELETE_STRICT commands to the OVS switch in order to delete a flow rule matching certain criteria (e.g., outport). What you expected to happen: The OVS switch receives the aforementioned command and deletes the flows that match the criteria. What actually happened: The matching criteria flows still appear in the OpenFlow tables. Although, I can see that apparently a flow was removed in the OVS database. The Open vSwitch version number: ovs-vswitchd (Open vSwitch) 2.15.90 The kernel version on which OVS is running: Linux version 4.15.0-140-generic (buildd@lgw01-amd64-054) (gcc version 7.5.0 (Ubuntu 7.5.0-3ubuntu1~18.04)) The distribution and version number of my OS: Ubuntu 18.04 The output of ovs-dptcl show: system@ovs-system: lookups: hit:5768706 missed:17566 lost:0 flows: 2 masks: hit:5955901 total:1 hit/pkt:1.03 port 0: ovs-system (internal) port 1: br2 (internal) port 2: ens4 port 3: ens5 port 4: ens6 port 5: ens7 The output of ovs-ofctl show : OFPT_FEATURES_REPLY (xid=0x2): dpid:faf67755c143 n_tables:254, n_buffers:0 capabilities: FLOW_STATS TABLE_STATS PORT_STATS QUEUE_STATS ARP_MATCH_IP actions: output enqueue set_vlan_vid set_vlan_pcp strip_vlan mod_dl_src mod_dl_dst mod_nw_src mod_nw_dst mod_nw_tos mod_tp_src mod_tp_dst 1(ens4): addr:fa:16:3e:8c:53:97 config: 0 state: 0 speed: 0 Mbps now, 0 Mbps max 2(ens5): addr:fa:16:3e:fd:c0:1a config: 0 state: 0 speed: 0 Mbps now, 0 Mbps max 3(ens6): addr:fa:16:3e:ba:d2:c0 config: 0 state: 0 speed: 0 Mbps now, 0 Mbps max 4(ens7): addr:fa:16:3e:67:bd:96 config: 0 state: 0 speed: 0 Mbps now, 0 Mbps max LOCAL(br2): addr:fa:f6:77:55:c1:43 config: 0 state: 0 speed: 0 Mbps now, 0 Mbps max OFPT_GET_CONFIG_REPLY (xid=0x4): frags=normal miss_send_len=0 A workaround is to establish an SSH connection to the host where OVS is running and execute ovs-ofctl del-flows . An automatic way to do it from the Ryu controller is using the Paramiko library to send remotely the command. We attach a screenshot where you can see the information in the OVS database at the moment the controller sends OFPFC_DELETE or OFPFC_DELETE_STRICT commands. Thanks you very much in advance. Kind regards, Alejandro Llorens -- El software de antivirus Avast ha analizado este correo electrónico en busca de virus. https://www.avast.com/antivirus ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
Re: [ovs-discuss] [ovn/ovs] OVS hardware offloading
Hello Ammad, On Wed, Jul 14, 2021 at 9:39 AM Ammad Syed wrote: > I am using OVN with openstack neutron on ubuntu 20.04. I want to enable > hardware offloading on ovs. I have used below command to to enable offload. > > ovs-vsctl set Open_vSwitch . other_config:hw-offload=true > > By reviewing below URL. > > https://www.openvswitch.org/support/ovscon2019/day2/0951-hw_offload_ovs_con_19-Oz-Mellanox.pdf > > I just want to enable kernel offload using tc. > > I am using Broadcom BCM57504 NetXtreme-E and below firmware and driver > details. > > driver: bnxt_en > version: 1.10.0 > firmware-version: 218.0.169.2/pkg 21.80.16.95 > expansion-rom-version: > bus-info: :63:00.0 > supports-statistics: yes > supports-test: yes > supports-eeprom-access: yes > supports-register-dump: no > supports-priv-flags: no Which kernel version are you using, and which version of OVN? > I have checked with ethtool -k with on NIC, hw-tc-offload: on is on. > > When I have enabled offloading. I have started seeing below errors. > > [ 1197.506839] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id > 0x1282 error 0x2 > [ 1197.507086] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: > cookie=0x9fe730da8000 error=-22 > [ 1197.516042] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id > 0x127f error 0x2 > [ 1197.525717] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: > cookie=0x9fe730da8000 error=-22 > [ 1227.522630] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id > 0x12f7 error 0x2 > [ 1227.531524] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: > cookie=0x9fe731a28800 error=-22 > [ 1227.540403] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id > 0x12f4 error 0x2 > [ 1227.548676] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: > cookie=0x9fe731a28800 error=-22 > [ 1257.563517] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id > 0x136c error 0x2 > [ 1257.572898] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: > cookie=0x9fe731a28800 error=-22 > [ 1257.583457] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id > 0x1369 error 0x2 > [ 1257.591647] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: > cookie=0x9fe731a28800 error=-22 > > Please advise how to fix it. I have not had the pleasure of testing this specific NIC myself, but from cursory view of kernel sources the message indicate that the driver is attempting an HWRM_CFA_FLOW_ALLOC operation and is getting a HWRM_ERR_CODE_INVALID_PARAMS error. My hunch would be there are some flow attributes in use that the driver/firmware does not support. If this is a 5.4 kernel it might be conntrack, you could try to disable port security for all networks and ports (make sure port-security is enabled in Neutron ML2 though, otherwise you would not be able to disable it properly). -- Frode Nordahl > > -- > Regards, > > > Ammad > ___ > discuss mailing list > disc...@openvswitch.org > https://mail.openvswitch.org/mailman/listinfo/ovs-discuss ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
[ovs-discuss] [ovn/ovs] OVS hardware offloading
Hi, I am using OVN with openstack neutron on ubuntu 20.04. I want to enable hardware offloading on ovs. I have used below command to to enable offload. ovs-vsctl set Open_vSwitch . other_config:hw-offload=true By reviewing below URL. https://www.openvswitch.org/support/ovscon2019/day2/0951-hw_offload_ovs_con_19-Oz-Mellanox.pdf I just want to enable kernel offload using tc. I am using Broadcom BCM57504 NetXtreme-E and below firmware and driver details. driver: bnxt_en version: 1.10.0 firmware-version: 218.0.169.2/pkg 21.80.16.95 expansion-rom-version: bus-info: :63:00.0 supports-statistics: yes supports-test: yes supports-eeprom-access: yes supports-register-dump: no supports-priv-flags: no I have checked with ethtool -k with on NIC, hw-tc-offload: on is on. When I have enabled offloading. I have started seeing below errors. [ 1197.506839] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x1282 error 0x2 [ 1197.507086] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe730da8000 error=-22 [ 1197.516042] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x127f error 0x2 [ 1197.525717] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe730da8000 error=-22 [ 1227.522630] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x12f7 error 0x2 [ 1227.531524] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1227.540403] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x12f4 error 0x2 [ 1227.548676] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1257.563517] bnxt_en :63:00.0 eth0-tor1: hwrm req_type 0x103 seq id 0x136c error 0x2 [ 1257.572898] bnxt_en :63:00.0 eth0-tor1: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 [ 1257.583457] bnxt_en :63:00.1 eth0-tor2: hwrm req_type 0x103 seq id 0x1369 error 0x2 [ 1257.591647] bnxt_en :63:00.1 eth0-tor2: Error: bnxt_tc_add_flow: cookie=0x9fe731a28800 error=-22 Please advise how to fix it. -- Regards, Ammad ___ discuss mailing list disc...@openvswitch.org https://mail.openvswitch.org/mailman/listinfo/ovs-discuss
