H3C S6805 Configuration Manual
Hide thumbs
Also See for S6805:
- Installation manual (61 pages) ,
- Troubleshooting manual (58 pages) ,
- Hardware information (24 pages)
Related Manuals for H3C S6805
Summary of Contents for H3C S6805
- Page 1 H3C S6805 & S6825 & S6850 & S9850 Switch Series Intelligent Lossless Network Configuration Guide New H3C Technologies Co., Ltd. http://www.h3c.com Software version: Release 6715 Document version: 6W100-20240402...
- Page 2 The information in this document is subject to change without notice. All contents in this document, including statements, information, and recommendations, are believed to be accurate, but they are presented without warranty of any kind, express or implied. H3C shall not be liable for technical or editorial errors or omissions contained herein.
-
Page 3: Command Conventions
This documentation is intended for: • Network planners. • Field technical support and servicing engineers. • Network administrators working with the S6805, S6825, S6850, or S9850 switch series. Conventions The following information describes the conventions used in the documentation. Command conventions Convention Description Boldface Bold text represents commands and keywords that you enter literally as shown. -
Page 4: Network Topology Icons
Symbols Convention Description An alert that calls attention to important information that if not understood or followed WARNING! can result in personal injury. An alert that calls attention to important information that if not understood or followed CAUTION: can result in data loss, data corruption, or damage to hardware or software. An alert that calls attention to essential information. -
Page 5: Documentation Feedback
Documentation feedback You can e-mail your comments about product documentation to info@h3c.com. We appreciate your comments. -
Page 6: Table Of Contents
Contents Configuring PFC ···························································································· 1 About PFC·························································································································································· 1 Restrictions: Hardware compatibility with PFC ·································································································· 1 Restrictions and guidelines ································································································································ 1 Configuring PFC on interfaces ··························································································································· 2 Configuring PFC on an Ethernet interface ································································································· 2 Setting PFC thresholds ······························································································································ 3 Configuring PFC deadlock detection·················································································································· 5 Configuring PFC deadlock prevention ···············································································································... -
Page 7: Configuring Pfc
Configuring PFC About PFC Priority-based flow control (PFC) provides a finer flow control mechanism to implement lossless packet transmission on Ethernet. PFC performs flow control for packets based on the 802.1p priorities carried in packets. As shown in Figure 1, PFC establishes eight virtual channels over an Ethernet link, each corresponding to an 802.1p priority. -
Page 8: Configuring Pfc On Interfaces
In an IRF network, follow these restrictions and guidelines: • For IRF and other protocols to operate correctly, as a best practice, do not enable PFC for 802.1p priority 0, 6, or 7. • To perform PFC on an IRF port, configure PFC on the IRF port and the IRF physical ports that are bound to the IRF port. -
Page 9: Setting Pfc Thresholds
interface interface-type interface-number Enable PFC on the Ethernet interface. priority-flow-control { auto | enable [ receive | send ] } By default, PFC is disabled. Enable PFC for 802.1p priorities. priority-flow-control no-drop dot1p dot1p-list By default, PFC is disabled for all 802.1p priorities. (Optional.) Set the pause time in PFC pause frames. - Page 10 Table 1 Default PFC thresholds Offset between the back PFC threshold Dynamic back Headroom pressure frame (right) pressure frame PFC reserved buffer stopping triggering threshold Interface type threshold threshold and threshold (below) triggering threshold 1-GE/10-GE 25-GE 40-GE 100-GE You must enable PFC for 802.1p priorities before setting the PFC thresholds. If you cancel PFC threshold settings on an interface, the PFC thresholds are restored to the state when only the command is executed.
-
Page 11: Configuring Pfc Deadlock Detection
priority-flow-control dot1p dot1p ingress-threshold-offset offset-number Table 1 for the default value. Set the PFC reserved threshold. priority-flow-control dot1p dot1p reserved-buffer reserved-number Table 1 for the default value. Configuring PFC deadlock detection About this task When packets carrying the specified 802.1p priority are transmitted in a loop, packets in the data buffer cannot be forwarded and PFC frames are repeatedly transmitted between devices. - Page 12 If you specify the keyword, the device automatically recovers the PFC pause-recover feature and PFC deadlock detection feature based on whether an interface receives PFC pause frames. If an interface is in PFC deadlock state and can still receive PFC pause frames when the ...
-
Page 13: Configuring Pfc Deadlock Prevention
Configuring PFC deadlock prevention About this task A device assigns an incoming packet to a queue with an 802.1p priority based on the DSCP value of the packet and the DSCP-802.1p priority map. When packets carrying the specified 802.1p priority are transmitted in a loop, each node on the path stops transmitting packets with the specified DSCP values. -
Page 14: Verifying And Maintaining Pfc
• The device fails and continuously sends PFC frames. In this case, you can set the early warning threshold for outgoing PFC packets. To monitor bidirectional PFC packets, you can set the early warning thresholds for incoming packets and outgoing packets separately. Restrictions and guidelines The number of PFC pause frames that an interface sends or receives is counted and the early warning threshold configuration takes effect only when PFC is enabled. - Page 15 Contents Configuring IPCC ··························································································· 1 About IPCC ························································································································································ 1 RoCEv2 packets ········································································································································ 1 IPCC mechanism ······································································································································· 2 Restrictions: Licensing requirements for IPCC ·································································································· 3 IPCC tasks at a glance······································································································································· 3 Enabling IPCC···················································································································································· 3 Enabling IPCC for a queue ································································································································ 4 Specifying an IPCC mode ··································································································································...
-
Page 16: Configuring Ipcc
Configuring IPCC About IPCC Intelligent Proactive Congestion Control (IPCC) is an improvement of Explicit Congestion Notification (ECN). It intelligently identifies the interface congestion state and proactively sends Congestion Notification Packets (CNPs) to control the RDMA over Converged Ethernet version 2 (RoCEv2) packet sending rate, ensuring low latency and high throughput. -
Page 17: Ipcc Mechanism
The Send, Write, and Read RoCEv2 packets are RoCEv2 data packets. Dest QP—Destination Queue Pair (QP), identifying an RoCEv2 flow. This field is similar to the destination port number. It is a key value for setting up an RoCEv2 flow table. IPCC mechanism IPCC is usually deployed in data center networks. -
Page 18: Restrictions: Licensing Requirements For Ipcc
Figure 2 IPCC mechanism Forwarding device Replicate RoCEv2 data packets, CM connection setup packets, or ACK packets and send the packets to CPU Obtain source IPs, destination The interface enabled with IPCC detects IPs, and destination QPs from the length of the queue enabled with RoCEv2 packets. -
Page 19: Enabling Ipcc For A Queue
Procedure Enter system view. system-view Enter interface view. interface interface-type interface-number Enable IPCC. ipcc enable By default, IPCC is disabled. Enabling IPCC for a queue About this task After you enable IPCC for a queue, the forwarding device monitors the length of the queue and calculates the number of CNPs to be proactively sent. -
Page 20: Ipcc Configuration Examples
Task Command display ipcc statistics [ interface Display statistics of CNPs that are proactively sent by the device. interface-type interface-number IPCC configuration examples Example: Configuring IPCC Network configuration As shown in Figure 3, the sender is deployed with storage services carried by RoCEv2. Enable IPCC on interface Twenty-FiveGigE1/0/2 on Device to control the RoCEv2 packet sending rate of the sender, ensuring low latency and high throughput for the RoCEv2 service. - Page 21 12.110.2.2 11.110.2.2 2090 ----------------------------------------------...
- Page 22 Contents Configuring iNOF ··························································································· 1 About iNOF ························································································································································ 1 Background ················································································································································ 1 Elements ···················································································································································· 1 Network types ············································································································································ 1 BFD-based link failure detection ················································································································ 3 Benefits ······················································································································································ 4 Protocols ···················································································································································· 5 License requirements ········································································································································· 5 Restrictions and guidelines ································································································································ 5 iNOF tasks at a glance ·······································································································································...
-
Page 23: Configuring Inof
Configuring iNOF About iNOF Background In Ethernet storage scenarios, the server side connects to the storage side over Ethernet. Traditionally, to achieve high throughput with packet loss during traffic transmission, the network server are configured to discover disk devices and establish long connections with them. If the server has not received any packets from a disk device for a long time, it considers the disk device as faulty and forwards storage traffic along the backup path. - Page 24 Figure 1 Directly connected network Default zone Server Disk Device iNOF Switch iNOF Switch Disk Device Server Zone 1 Cross-switch network As shown in Figure 2, the hosts in the same iNOF zone can connect to different iNOF switches. These switches exchange dynamic host information. This iNOF network type is applicable to large-scale data centers with multiple hosts from different organizations or departments.
-
Page 25: Bfd-Based Link Failure Detection
Figure 2 Cross-switch network Zone 1 Default zone …… …… Device C (Client 2) Device B Device D (Reflector 1) (Reflector 2) Device E Device A (Client 3) (Client 1) … … Zone 2 … … …… Zone 1 Default zone iNOF Link Common Link NOTE:... -
Page 26: Benefits
BGP-BFD collaboration. a. After the local BGP device establishes an IBGP peer session to another BGP device, the local device delivers information about the IBGP peer to the BFD module. IBGP peer information includes destination address and source address. b. On receipt of IBGP peer information, the BFD module automatically establishes a BFD session to the IBGP peer. -
Page 27: Protocols
• High data security There are various encryption methods for BGP sessions, including MD5 authentication, GTSM, and keychain authentication. These encryption methods enhance the data security among iNOF switches. Protocols The iNOF switch selects one of the following protocols to exchange packets with hosts after negotiating with them. -
Page 28: Inof Tasks At A Glance In Cross-Switch Inof Networks
iNOF tasks at a glance in cross-switch iNOF networks Configuring iNOF clients Configuring an iNOF client Enabling iNOF Configuring iNOF reflectors Configuring an iNOF reflector (Optional.) Enabling automatic adding of free hosts to the default iNOF zone ... -
Page 29: Configuring An Inof Reflector
If no loopback interface IP address is available, BGP uses the highest physical interface IP address as the route ID regardless of the interface status. Enter BGP instance view. bgp as-number [ instance instance-name ] By default, BGP is disabled and no BGP instances exist. (Optional.) Configure a router ID for the BGP instance. -
Page 30: Enabling Inof
(Optional.) Configure a router ID for the BGP instance. router-id router-id By default, no router ID is configured for a BGP instance, and the BGP instance uses the global router ID configured by the command in system view. router-id (Optional.) Configure BFD for the link to an IPv4 BGP peer. peer ipv4-address [ mask-length ] bfd [ echo | multi-hop | single-hop ] By default, BFD is disabled. -
Page 31: Creating An Inof Zone
Restrictions and guidelines If an iNOF reflector exists in the network, you must enable this feature on the iNOF reflector. For an iNOF client on such a network, whether to add free hosts to the default zone depends on the configuration of this feature for the iNOF reflector. -
Page 32: Display And Maintenance Commands For Inof
Restrictions and guidelines If an iNOF reflector exists in the network, you must enable this feature on the iNOF reflector. For an iNOF client on such a network, the status of iNOF hard zoning depends on the configuration of this feature for the iNOF reflector. -
Page 33: Inof Configuration Examples
iNOF configuration examples Directly-connected iNOF network configuration examples Network configuration As shown in Figure 3, the network server and the disk device are connected through Device A and Device B. When the disk device fails, the server can detect the failure promptly. Figure 3 Directly-connected iNOF network diagram Zone 1 Disk device... -
Page 34: Cross-Switch Inof Network Configuration Examples
[DeviceA-GigabitEthernet1/0/2] quit Configure Device B. Details are not shown. The configuration of Device B is the same as that of Device A. Verify the configuration # Display information about user-defined iNOF zones. [DeviceA] display inof configuration zone Total Zone number: 2 iNOF Default-Zone: Enable ZoneName Host... - Page 35 # Assign IP addresses to VLAN-interface 10 and VLAN-interface 30. These interfaces are used for BGP connections. <DeviceA> system-view [DeviceA] vlan 10 30 [DeviceA] interface vlan 10 [DeviceA-Vlan-interface10] ip address 10.1.1.2 24 [DeviceA-Vlan-interface10] quit [DeviceA] interface vlan 30 [DeviceA-Vlan-interface30] ip address 30.1.1.2 24 [DeviceA-Vlan-interface30] quit # Specify Device B and Device C as BGP peers of Device A.
- Page 36 [DeviceB-bgp-default] peer 100.1.1.2 as-number 100 [DeviceB-bgp-default] peer 100.1.1.2 bfd [DeviceB-bgp-default] address-family inof [DeviceB-bgp-default-inof] role reflector [DeviceB-bgp-default-inof] peer 10.1.1.2 enable [DeviceB-bgp-default-inof] reflector cluster-id 88.88.88.88 [DeviceB-bgp-default-inof] peer 10.1.1.2 reflect-client [DeviceB-bgp-default-inof] peer 20.1.1.2 enable [DeviceB-bgp-default-inof] peer 20.1.1.2 reflect-client [DeviceB-bgp-default-inof] peer 100.1.1.2 enable [DeviceB-bgp-default-inof] peer 100.1.1.2 reflect-client [DeviceB-bgp-default-inof] quit [DeviceB-bgp-default] quit # Enable LLDP globally.
- Page 37 [DeviceC-bgp-default] peer 100.1.1.1 bfd [DeviceC-bgp-default] address-family inof [DeviceC-bgp-default-inof] role reflector [DeviceC-bgp-default-inof] reflector cluster-id 88.88.88.88 [DeviceC-bgp-default-inof] peer 30.1.1.2 enable [DeviceC-bgp-default-inof] peer 30.1.1.2 reflect-client [DeviceC-bgp-default-inof] peer 40.1.1.2 enable [DeviceC-bgp-default-inof] peer 40.1.1.2 reflect-client [DeviceC-bgp-default-inof] peer 100.1.1.1 enable [DeviceC-bgp-default-inof] peer 100.1.1.1 reflect-client [DeviceC-bgp-default-inof] quit [DeviceC-bgp-default] quit # Enable LLDP globally.
- Page 38 [DeviceD-inof] quit Verify the configuration # Display iNOF reflector information on Device A. <DeviceA> display inof reflector Index ReflectorIP DefaultZone HardZoning 10.1.1.1 Enabled Disabled 10.1.1.2 Enabled Disabled # Display iNOF reflector information on Device B. <DeviceB> display inof reflector Index ReflectorIP DefaultZone HardZoning...
Need help?
Do you have a question about the S6805 and is the answer not in the manual?
Questions and answers