Connectivity Guide
Table Of Contents
- OS10 Enterprise Edition User Guide Release 10.4.2.0
- Getting Started
- Supported Hardware
- Download OS10 image and license
- Installation using ONIE
- Log into OS10
- Install OS10 license
- Zero-touch deployment
- Remote access
- Upgrade OS10
- CLI Basics
- User accounts
- Key CLI features
- CLI command modes
- CLI command hierarchy
- CLI command categories
- CONFIGURATION Mode
- Command help
- Check device status
- Candidate configuration
- Change to transaction-based configuration mode
- Copy running configuration
- Restore startup configuration
- Reload system image
- Filter show commands
- Alias command
- Batch mode
- Linux shell commands
- SSH commands
- OS9 environment commands
- Common commands
- alias
- alias (multi-line)
- batch
- boot
- commit
- configure
- copy
- default (alias)
- delete
- description (alias)
- dir
- discard
- do
- feature config-os9-style
- exit
- license
- line (alias)
- lock
- management route
- move
- no
- reload
- show alias
- show boot
- show candidate-configuration
- show environment
- show inventory
- show ip management-route
- show ipv6 management-route
- show license status
- show running-configuration
- show startup-configuration
- show system
- show version
- start
- system
- system identifier
- terminal
- traceroute
- unlock
- write
- Interfaces
- Ethernet interfaces
- Unified port groups
- L2 mode configuration
- L3 mode configuration
- Fibre Channel interfaces
- Management interface
- VLAN interfaces
- User-configured default VLAN
- VLAN scale profile
- Loopback interfaces
- Port-channel interfaces
- Configure interface ranges
- Switch-port profiles
- Configure breakout mode
- Breakout auto-configuration
- Forward error correction
- Energy-efficient Ethernet
- View interface configuration
- Interface commands
- channel-group
- default vlan-id
- description (Interface)
- duplex
- feature auto-breakout
- fec
- interface breakout
- interface ethernet
- interface loopback
- interface mgmt
- interface null
- interface port-channel
- interface range
- interface vlan
- link-bundle-utilization
- mode
- mode l3
- mtu
- port-group
- scale-profile vlan
- show interface
- show inventory media
- show link-bundle-utilization
- show port-channel summary
- show port-group
- show switch-port-profile
- show vlan
- shutdown
- speed (Fibre Channel)
- speed (Management)
- switch-port-profile
- switchport access vlan
- switchport mode
- switchport trunk allowed vlan
- Fibre Channel
- Layer 2
- 802.1X
- Link Aggregation Control Protocol
- Link Layer Discovery Protocol
- Protocol data units
- Optional TLVs
- Organizationally-specific TLVs
- Media endpoint discovery
- Network connectivity device
- LLDP-MED capabilities TLV
- Network policies TLVs
- Define network policies
- Packet timer values
- Disable and re-enable LLDP
- Disable and re-enable LLDP on management ports
- Advertise TLVs
- Network policy advertisement
- Fast start repeat count
- View LLDP configuration
- Adjacent agent advertisements
- Time to live
- LLDP commands
- Media Access Control
- Multiple Spanning-Tree
- Rapid per-VLAN spanning-tree plus
- Rapid Spanning-Tree Protocol
- Virtual LANs
- Port monitoring
- Layer 3
- Virtual routing and forwarding
- Bidirectional Forwarding Detection
- Border Gateway Protocol
- Sessions and peers
- Route reflectors
- Multiprotocol BGP
- Attributes
- Selection criteria
- Weight and local preference
- Multiexit discriminators
- Origin
- AS path and next-hop
- Best path selection
- More path support
- Advertise cost
- 4-Byte AS numbers
- AS number migration
- Configure Border Gateway Protocol
- Enable BGP
- Configure Dual Stack
- Configure administrative distance
- Peer templates
- Neighbor fall-over
- Configure password
- Fast external fallover
- Passive peering
- Local AS
- AS number limit
- Redistribute routes
- Additional paths
- MED attributes
- Local preference attribute
- Weight attribute
- Enable multipath
- Route-map filters
- Route reflector clusters
- Aggregate routes
- Confederations
- Route dampening
- Timers
- Neighbor soft-reconfiguration
- BGP commands
- Equal cost multi-path
- IPv4 routing
- IPv6 routing
- Internet Group Management Protocol
- Multicast Listener Discovery Protocol
- Open shortest path first
- Object tracking manager
- Policy-based routing
- Virtual Router Redundancy Protocol
- VXLAN
- VXLAN concepts
- VXLAN as NVO solution
- Configure VXLAN
- VXLAN commands
- member-interface
- nve
- remote-vtep
- show nve remote-vtep
- show nve remote-vtep counters
- show nve vxlan-vni
- show virtual-network
- show virtual-network counters
- show virtual-network interface counters
- show virtual-network interface
- show virtual-network vlan
- show vlan (virtual network)
- source-interface loopback
- virtual-network
- virtual-network untagged-vlan
- vxlan-vni
- VXLAN MAC commands
- clear mac address-table dynamic nve remote-vtep
- clear mac address-table dynamic virtual-network
- show mac address-table count extended
- show mac address-table count nve
- show mac address-table count virtual-network
- show mac address-table extended
- show mac address-table nve
- show mac address-table virtual-network
- Example: VXLAN with static VTEP
- BGP EVPN for VXLAN
- UFT modes
- System management
- Dynamic Host Configuration Protocol
- Network Time Protocol
- System clock
- System banners
- User session management
- Telnet server
- Security
- User re-authentication
- Password strength
- Role-based access control
- Assign user role
- RADIUS authentication
- TACACS+ authentication
- TACACS+ unknown or missing user role
- SSH server
- Virtual terminal line
- Enable AAA accounting
- Enable user lockout
- Limit concurrent login sessions
- Enable login statistics
- Security commands
- Simple Network Management Protocol
- OS10 image upgrade
- OpenFlow
- Access Control Lists
- IP ACLs
- MAC ACLs
- Control-plane ACLs
- IP fragment handling
- L3 ACL rules
- Assign sequence number to filter
- Delete ACL rule
- L2 and L3 ACLs
- Assign and apply ACL filters
- Ingress ACL filters
- Egress ACL filters
- Clear access-list counters
- IP prefix-lists
- Route-maps
- Match routes
- Set conditions
- Continue clause
- ACL flow-based monitoring
- Enable flow-based monitoring
- ACL table profiles
- View ACL table utilization report
- ACL commands
- acl-table-profile
- clear ip access-list counters
- clear ipv6 access-list counters
- clear mac access-list counters
- deny
- deny (IPv6)
- deny (MAC)
- deny icmp
- deny icmp (IPv6)
- deny ip
- deny ipv6
- deny tcp
- deny tcp (IPv6)
- deny udp
- deny udp (IPv6)
- description
- hardware acl-table-profile
- ingress app-group
- ip access-group
- ip access-list
- ip as-path access-list
- ip community-list standard deny
- ip community–list standard permit
- ip extcommunity-list standard deny
- ip extcommunity-list standard permit
- ip prefix-list description
- ip prefix-list deny
- ip prefix-list permit
- ip prefix-list seq deny
- ip prefix-list seq permit
- ipv6 access-group
- ipv6 access-list
- ipv6 prefix-list deny
- ipv6 prefix-list description
- ipv6 prefix-list permit
- ipv6 prefix-list seq deny
- ipv6 prefix-list seq permit
- mac access-group
- mac access-list
- permit
- permit (IPv6)
- permit (MAC)
- permit icmp
- permit icmp (IPv6)
- permit ip
- permit ipv6
- permit tcp
- permit tcp (IPv6)
- permit udp
- permit udp (IPv6)
- remark
- seq deny
- seq deny (IPv6)
- seq deny (MAC)
- seq deny icmp
- seq deny icmp (IPv6)
- seq deny ip
- seq deny ipv6
- seq deny tcp
- seq deny tcp (IPv6)
- seq deny udp
- seq deny udp (IPv6)
- seq permit
- seq permit (IPv6)
- seq permit (MAC)
- seq permit icmp
- seq permit icmp (IPv6)
- seq permit ip
- seq permit ipv6
- seq permit tcp
- seq permit tcp (IPv6)
- seq permit udp
- seq permit udp (IPv6)
- show access-group
- show access-lists
- show acl-table-profile
- show acl-table-usage detail
- show ip as-path-access-list
- show ip community-list
- show ip extcommunity-list
- show ip prefix-list
- Route-map commands
- continue
- match as-path
- match community
- match extcommunity
- match interface
- match ip address
- match ip next-hop
- match ipv6 address
- match ipv6 next-hop
- match metric
- match origin
- match route-type
- match tag
- route-map
- set comm-list add
- set comm-list delete
- set community
- set extcomm-list add
- set extcomm-list delete
- set extcommunity
- set local-preference
- set metric
- set metric-type
- set next-hop
- set origin
- set tag
- set weight
- show route-map
- Quality of service
- Configure quality of service
- Ingress traffic classification
- Egress traffic classification
- Policing traffic
- Mark Traffic
- Color traffic
- Modify packet fields
- Shaping traffic
- Bandwidth allocation
- Strict priority queuing
- Buffer management
- Congestion avoidance
- Storm control
- RoCE for faster access and lossless connectivity
- QoS commands
- bandwidth
- class
- class-map
- clear interface
- clear qos statistics
- clear qos statistics type
- control-plane
- control-plane-buffer-size
- flowcontrol
- match
- match cos
- match dscp
- match precedence
- match queue
- match vlan
- mtu
- pause
- pfc-cos
- pfc-max-buffer-size
- pfc-shared-buffer-size
- pfc-shared-headroom-buffer-size
- police
- policy-map
- priority
- priority-flow-control mode
- qos-group dot1p
- qos-group dscp
- queue-limit
- queue bandwidth
- queue qos-group
- random-detect (interface)
- random-detect (queue)
- random-detect color
- random-detect ecn
- random-detect ecn
- random-detect pool
- random-detect weight
- service-policy
- set cos
- set dscp
- set qos-group
- shape
- show class-map
- show control-plane buffers
- show control-plane buffer-stats
- show control-plane info
- show control-plane statistics
- show interface priority-flow-control
- show qos interface
- show policy-map
- show qos control-plane
- show qos egress bufffers interface
- show egress buffer-stats interface
- show qos ingress buffers interface
- show ingress buffer-stats interface
- show queuing statistics
- show qos system
- show qos system buffers
- show qos maps
- show qos wred-profile
- system qos
- trust-map
- trust dot1p-map
- trust dscp-map
- qos-map traffic-class
- trust-map
- wred
- Virtual Link Trunking
- Uplink Failure Detection
- Converged data center services
- sFlow
- RESTCONF API
- Troubleshoot OS10
- Support resources
Table 51. Default settings for PFC
Speed 10G 25G 40G 50G 100G
Default reserved buer for S4000, S4048–ON, S6010–ON 9KB NA 9KB NA NA
Default reserved buer for S41xx, Z9100–ON 9KB 9KB 18KB 18KB 36KB
Default Xo threshold 36KB 45KB 75KB 91KB 142KB
Default Xon threshold 9KB 9KB 9KB 9KB 9KB
Default dynamic share buer threshold (alpha value) 9KB 9KB 9KB 9KB 9KB
NOTE: The supported speed varies for dierent platforms. After the reserved buers are used, each PFC starts consuming
shared buers from the lossless pool with the alpha value determining the threshold.
You can override the default priority group settings when you enable LLFC or PFC.
1 Create a network-qos type class-map to match the trac classes. For LLFC, match all the trac classes from 0 to 7. For PFC, match
the required trac class.
OS10(config)# class-map type network-qos tc
OS10 (config-cmap-nqos)# match qos-group 0-7
2 Create network-qos type policy-map to dene the actions for trac classes, such as a buer conguration and threshold.
OS10(config)# policy-map type network-qos buffer
OS10(config-pmap-network-qos)# class tc
OS10 (config-pmap-c-nqos)# pause buffer-size 300 pause-threshold 200 resume-threshold 100
OS10 (config-pmap-c-nqos)# queue-limit thresh-mode dynamic 5
Congure egress buer
All port queues are allocated with reserved buers. When the reserved buers are consumed, each queue starts using the shared buers
from the default pool.
The reserved buer per queue is 1664 bytes for the speed of 10G, 25G, 40G, 50G, and 100G. The default dynamic shared buer threshold
is 8.
1 Create a queuing type class-map to match the queue.
OS10(config)# class-map type queuing q1
OS10(config-cmap-queuing)# match queue 1
2 Create a queuing type policy-map to dene the actions for queues, such as a buer conguration and threshold.
OS10(config)# policy-map type queuing q-buffer
OS10(config-pmap-queuing)# class q1
OS10(config-pmap-c-que)# queue-limit queue-len 200 thresh-mode dynamic 5
Congestion avoidance
Congestion avoidance anticipates and takes necessary actions to avoid congestion. The following mechanisms avoid congestion:
• Tail drop—Packets are buered at trac queues. When the buers are exhausted or reach the congured threshold, excess packets
drop. By default, OS10 uses tail drop for congestion avoidance.
• Random early detection (RED)—In tail drop, dierent ows are not considered in buer utilization. When multiple hosts start
retransmission, tail drop causes TCP global re-synchronization. Instead of waiting for the queue to get lled up completely, RED starts
dropping excess packets with a certain drop-probability when the average queue length exceeds the congured minimum threshold.
The early drop ensures that only some of TCP sources slow down, which avoids global TCP re-synchronization.
• Weighted random early detection (WRED)—This allows dierent drop-probabilities and thresholds for each color — red, yellow, green
— of trac. You can congure the drop characteristics for three dierent ows by assigning the colors to the ow. Assign colors to a
particular ow or trac using various methods, such as ingress policing, qos input policy-maps, and so on.
792
Quality of service