Users Guide
Table Of Contents
- OS10 Enterprise Edition User Guide Release 10.3.2E(R2)
- Getting Started
- Download OS10 image and license
- Installation
- Log into OS10
- Install OS10 license
- 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
- Back up or restore configuration
- Reload system image
- Filter show commands
- Alias command
- Batch mode commands
- Linux shell commands
- SSH commands
- OS9 environment commands
- Common commands
- alias
- batch
- boot
- commit
- configure
- copy
- delete
- dir
- discard
- do
- feature config-os9-style
- exit
- license
- 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
- L2 mode configuration
- L3 mode configuration
- Management interface
- VLAN interfaces
- Loopback interfaces
- Port-channel interfaces
- Create port-channel
- Add port member
- Minimum links
- Assign Port Channel IP Address
- Remove or disable port-channel
- Load balance traffic
- Change hash algorithm
- Configure interface ranges
- Forward error correction
- View interface configuration
- Interface commands
- channel-group
- description (Interface)
- duplex
- fec
- interface breakout
- interface ethernet
- interface loopback
- interface mgmt
- interface null
- interface port-channel
- interface range
- interface vlan
- link-bundle-utilization
- mgmt
- mtu
- show interface
- show link-bundle-utilization
- show port-channel summary
- show vlan
- shutdown
- speed (Management)
- switchport access vlan
- switchport mode
- switchport trunk allowed vlan
- 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
- 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 protocol
- Rapid per-VLAN spanning-tree plus
- Rapid spanning-tree protocol
- Virtual LANs
- Port monitoring
- Layer 3
- 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
- Peer templates
- Neighbor fall-over
- 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
- Open shortest path first
- Object tracking manager
- Policy-based routing
- Virtual router redundancy protocol
- Border gateway protocol
- System management
- Access Control Lists
- IP ACLs
- MAC ACLs
- IP fragment handling
- L3 ACL rules
- Assign sequence number to filter
- 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 commands
- 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
- ip access-group
- ip access-list
- ip as-path deny
- ip as-path permit
- 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 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 delete
- set community
- 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
- Class-map configuration
- Policy-map configuration
- Ingress traffic classification
- Queue selection
- Strict priority queuing
- Class of service or dot1p classification
- Mark traffic
- Traffic metering
- Bandwidth allocation
- Service-policy rate-shaping
- Policy-based rate-policing
- Storm control
- Control-plane policing
- Queue management
- Verify configuration
- Egress queue statistics
- QoS commands
- bandwidth
- class
- class-map
- clear interface
- clear qos statistics
- clear qos statistics type
- control-plane
- flowcontrol
- match
- match cos
- match dscp
- match precedence
- match queue
- match vlan
- pause
- police
- policy-map
- priority
- qos-group dot1p
- qos-group dscp
- queue qos-group
- random-detect
- service-policy
- set cos
- set dscp
- set qos-group
- shape
- show class-map
- show control-plane info
- show control-plane statistics
- 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 qos system
- show qos system buffers
- show qos maps
- system qos
- trust
- trust dot1p-map
- trust dscp-map
- qos-map traffic-class
- trust-map
- Virtual link trunking
- Converged data center services
- sFlow
- Troubleshoot OS10
- Support resources
PFC conguration notes
• PFC is supported for 802.1p priority trac (dot1p 0 to 7). FCoE trac traditionally uses dot1p priority 3 — iSCSI storage trac uses
dot1p priority 4.
• Congure PFC for ingress trac by using network-qos class and policy maps (see Quality of Service). The queues used for PFC-
enabled trac are treated as lossless queues. Congure the same network-qos policy map on all PFC-enabled ports. Congure required
bandwidth for lossless trac using ETS queuing (output) policies on egress interfaces.
• In a network-qos policy-class map, use commands to generate PFC pause frames for matching class-map priorities:
– Send pause frames for matching class-map trac during congestion (pause command).
– (Optional) Enter user-dened values for the reserved ingress buer-size of PFC class-map trac, and the thresholds used to send
XOFF and XON pause frames (pause [buffer-size kilobytes pause-threshold kilobytes resume-
threshold
kilobytes]command).
– Congure the matching dot1p values used to send pause frames (pfc-cos command).
• By default, all ingress trac is handled by the lossy ingress buer. When you enable PFC, dot1p ingress trac competes for shared
buers in the lossless pool instead of the shared lossy pool. The number of lossless queues supported on an interface depends on the
amount of available free memory in the lossy pool.
• Use the priority-flow-control mode on command to enable PFC for FCoE and iSCSI trac (example, priority 3 and 4).
• Enable DCBX on interfaces to detect and auto-congure PFC/ETS parameters from peers.
• PFC and 802.3x link-level ow control (LLFC) are disabled by default on an interface. You cannot enable PFC and LLFC at the same
time. LLFC ensures lossy trac in best-eort transmission. Enable PFC to enable guarantee lossless FCoE and iSCSI trac. PFC
manages buer congestion by pausing specied ingress dot1p trac; LLFC pauses all data transmission on an interface. To enable
LLFC, enter the flowcontrol [receive | transmit] [on | off] command.
• SYSTEM-QOS mode applies a service policy globally on all interfaces:
– Create and apply a 1-to-1 802.1p-priority-to-trac-class mapping on an interface or all interfaces in INTERFACE or SYSTEM-QOS
mode
– Create and apply a 1-to-1 trac-class-to-queue mapping on an interface or all interfaces in INTERFACE or SYSTEM-QOS mode
The S5148F-ON platform has the following limitations:
• You cannot congure PFC priority 0 as a lossless priority.
• You cannot map multiple priorities to same queue.
• Whenever LLFC is enabled on an interface, Rx PFC frames are honored. Also, whenever PFC is enabled on an interface, Rx Pause
frames are honored. With respect to statistics, Rx Pause statistics in the hardware includes the Rx PFC frames too.
Congure dot1p priority to trac class mapping
Decide if you want to use the default 802.1p priority-to-trac class (qos-group) mapping or congure a new map. By default, the qos
class-trust class map is applied to ingress trac. Class-trust is a reserved class name. The class-trust class instructs OS10 interfaces to
honor dot1p or DSCP trac.
Dot1p Priority : 0 1 2 3 4 5 6 7
Traffic Class : 1 0 2 3 4 5 6 7
Converged data center services
501