Step 1 |
Server#
scope chassis
|
Enters the
chassis command mode.
|
Step 2 |
Server /chassis
#
show
adapter
|
(Optional)
Displays the available adapter devices.
|
Step 3 |
Server /chassis
#
scope adapter
index
|
Enters the
command mode for the adapter card at the PCI slot number specified by
index .
Note
|
The server
must be powered on before you can view or change adapter settings.
|
|
Step 4 |
Server
/chassis/adapter #
scope
host-eth-if
{eth0 |
eth1 |
name}
|
Enters the host
Ethernet interface command mode for the specified vNIC.
|
Step 5 |
Server
/chassis/adapter/host-eth-if #
set
mtu
mtu-value
|
Specifies the
maximum transmission unit (MTU) or packet size that the vNIC accepts. Valid MTU
values are 1500 to 9000 bytes; the default is 1500.
|
Step 6 |
Server
/chassis/adapter/host-eth-if #
set
uplink
{0
|
1}
|
Specifies the
uplink port associated with this vNIC. All traffic for this vNIC goes through
this uplink port.
|
Step 7 |
Server
/chassis/adapter/host-eth-if #
set
mac-addr
mac-addr
|
Specifies a MAC
address for the vNIC in the form hh:hh:hh:hh:hh:hh or hhhh:hhhh:hhhh.
|
Step 8 |
Server
/chassis/adapter/host-eth-if #
set
cos
cos-value
|
Specifies the
class of service (CoS) value to be marked on received packets unless the vNIC
is configured to trust host CoS. Valid CoS values are 0 to 6; the default is 0.
Higher values indicate more important traffic.
Note
|
-
You must set the
COS value to 5 for the RDMA enabled
interfaces.
-
If NIV is enabled, this setting is determined by the switch,
and the command is ignored.
|
|
Step 9 |
Server
/chassis/adapter/host-eth-if #
set
trust-host-cos {disable
|
enable}
|
Specifies whether the vNIC will trust host CoS or will remark packets. The behavior is as follows:
|
Step 10 |
Server
/chassis/adapter/host-eth-if #
set
order
{any
|
0-99}
|
Specifies the
relative order of this device for PCI bus device number assignment; the default
is any.
|
Step 11 |
Server
/chassis/adapter/host-eth-if #
set
vlan
{none
|
vlan-id}
|
Specifies the
default VLAN for this vNIC. Valid VLAN numbers are 1 to 4094; the default is
none.
Note
|
If NIV is
enabled, this setting is determined by the switch, and the command is ignored.
|
|
Step 12 |
Server
/chassis/adapter/host-eth-if #
set
vlan-mode {access
|
trunk}
|
Specifies the VLAN mode for the vNIC. The modes are as follows:
-
access
—The vNIC belongs to only one VLAN. When the VLAN is set to access mode, any frame received from the specified default VLAN
(1-4094) that is received from the switch with a TAG removes that TAG when it is sent to the host OS through the vNIC.
-
trunk
—The vNIC can belong to more than one VLAN. This is the default.
Note
|
If NIV is
enabled, this setting is determined by the switch, and the command is ignored.
|
|
Step 13 |
Server
/chassis/adapter/host-eth-if #
set
rate-limit {off
|
rate}
|
Specifies a maximum data rate for the vNIC. The range is 1 to 10000 Mbps; the default is off.
For VIC 13xx controllers, you can enter an integer between 1 and 40,000.
For VIC 1455 and 1457 controllers:
-
If the adapter is connected to 25 Gbps link on a switch, then you can enter an integer between 1 to 25,000 Mbps.
-
If the adapter is connected to 10 Gbps link on a switch, then you can enter an integer between 1 to 10,000 Mbps.
For VIC 1495 and 1497 controllers:
-
If the adapter is connected to 40 Gbps link on a switch, then you can enter an integer between 1 to 40,000 Mbps.
-
If the adapter is connected to 100 Gbps link on a switch, then you can enter an integer between 1 to 100,000 Mbps.
Note
|
If NIV is enabled, this setting is determined by the switch, and the command is ignored.
|
|
Step 14 |
Server
/chassis/adapter/host-eth-if #
set
boot
{disable
|
enable}
|
Specifies whether the vNIC can be used to perform a PXE boot. Default value is set to disable for the default vNICs and user-created
vNICs.
|
Step 15 |
Server
/chassis/adapter/host-eth-if #
set
channel-number
number
|
If NIV mode is
enabled for the adapter, select the channel number that will be assigned to
this vNIC. The range is 1 to 1000.
|
Step 16 |
Server
/chassis/adapter/host-eth-if #
set
port-profile
name
|
If NIV mode is
enabled for the adapter, select the port profile that should be associated with
the vNIC.
Note
|
The
name must be
a port profile defined on the switch to which this server is connected.
|
|
Step 17 |
Server
/chassis/adapter/host-eth-if #
set
uplink-failover {disable
|
enable}
|
If NIV mode is
enabled for the adapter, enable this setting if traffic on this vNIC should
fail over to the secondary interface if there are communication problems.
|
Step 18 |
Server
/chassis/adapter/host-eth-if #
set
uplink-failback-timeout
seconds
|
After a vNIC
has started using its secondary interface, this setting controls how long the
primary interface must be available before the system resumes using the primary
interface for the vNIC.
Enter a number
of
seconds
between 0 and 600.
|
Step 19 |
Server /chassis/adapter/host-eth-if #
set
vmq {disabled
|
enabled}
|
Enables or disables Virtual Machine Queue (VMQ) for this adapter.
|
Step 20 |
Server /chassis/adapter/host-eth-if #
set
multi-queue {disabled
|
enabled}
|
Enables or disables the multi queue option for this adapter and allows you to set the following multi queue parameters:
-
mq-rq-count —The number of receive queue resources to allocate. Enter an integer between 1 and 1000.
-
mq-wq-count —The number of transmit queue resources to allocate. Enter an integer between 1 and 1000.
-
mq-cq-count —The number of completion queue resources to allocate. In general, the number of completion queue resources you should allocate
is equal to the number of transmit queue resources plus the number of receive queue resources. Enter an integer between 1
and 2000.
Note
|
-
Multi queue is supported only on C-Series servers with 14xx adapters.
-
VMQ must be in enabled state to enable this option.
-
When you enable this option on one of the vNICs, configuring only VMQ (without choosing multi-queue) on other vNICs is not
supported.
-
When this option is enabled usNIC configuration will be disabled.
|
|
Step 21 |
Server
/chassis/adapter/host-eth-if #
set
arfs
{disable
|
enable}
|
Enables or
disables Accelerated Receive Flow steering (aRFS) for this adapter.
|
Step 22 |
Server
/chassis/adapter/host-eth-if #
scope
interrupt
|
Enters the
interrupt command mode.
|
Step 23 |
Server
/chassis/adapter/host-eth-if/interrupt #
set
interrupt-count
count
|
Specifies the
number of interrupt resources. The range is 1 to 514; the default is 8. In
general, you should allocate one interrupt resource for each completion queue.
|
Step 24 |
Server
/chassis/adapter/host-eth-if/interrupt #
set
coalescing-time
usec
|
The time to
wait between interrupts or the idle period that must be encountered before an
interrupt is sent.
The range is 1
to 65535 microseconds; the default is 125. To turn off coalescing, enter 0
(zero).
|
Step 25 |
Server
/chassis/adapter/host-eth-if/interrupt #
set
coalescing-type {idle
|
min}
|
The coalescing
types are as follows:
-
idle
—The system does not send an interrupt until there is a
period of no activity lasting as least as long as the time specified in the
coalescing time configuration.
-
min
—The system waits for the time specified in the
coalescing time configuration before sending another interrupt event. This is
the default.
|
Step 26 |
Server
/chassis/adapter/host-eth-if/interrupt #
set
interrupt-mode {intx
|
msi
|
msix}
|
Specifies the Ethernet interrupt mode. The modes are as follows:
-
intx
—Line-based interrupt (PCI INTx)
-
msi
—Message-Signaled Interrupt (MSI)
-
msix
—Message Signaled Interrupts with the optional extension (MSI-X). This is the recommended and default option.
|
Step 27 |
Server
/chassis/adapter/host-eth-if/interrupt #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 28 |
Server
/chassis/adapter/host-eth-if #
scope
recv-queue
|
Enters receive
queue command mode.
|
Step 29 |
Server
/chassis/adapter/host-eth-if/recv-queue #
set
rq-count
count
|
The number of
receive queue resources to allocate. The range is 1 to 256; the default is 4.
|
Step 30 |
Server
/chassis/adapter/host-eth-if/recv-queue #
set
rq-ring-size
size
|
The number of
descriptors in the receive queue. The range is 64 to 4094; the default is 512.
|
Step 31 |
Server
/chassis/adapter/host-eth-if/recv-queue #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 32 |
Server
/chassis/adapter/host-eth-if #
scope
trans-queue
|
Enters
transmit queue command mode.
|
Step 33 |
Server
/chassis/adapter/host-eth-if/trans-queue #
set
wq-count
count
|
The number of
transmit queue resources to allocate. The range is 1 to 256; the default is 1.
|
Step 34 |
Server
/chassis/adapter/host-eth-if/trans-queue #
set
wq-ring-size
size
|
The number of
descriptors in the transmit queue. The range is 64 to 4094; the default is 256.
|
Step 35 |
Server
/chassis/adapter/host-eth-if/trans-queue #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 36 |
Server
/chassis/adapter/host-eth-if #
scope
comp-queue
|
Enters
completion queue command mode.
|
Step 37 |
Server
/chassis/adapter/host-eth-if/comp-queue #
set
cq-count
count
|
The number of
completion queue resources to allocate. The range is 1 to 512; the default is
5.
In general,
the number of completion queues equals the number of transmit queues plus the
number of receive queues.
|
Step 38 |
Server
/chassis/adapter/host-eth-if/comp-queue #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 39 |
Server /chassis/adapter/host-eth-if/ #
set
rdma_mr
number
|
Sets the number of memory regions to be used per adapter. The values range from 4096 to 524288.
|
Step 40 |
Server /chassis/adapter/host-eth-if/ #
set
rdma_qp
number
|
Sets the number of queue pairs to be used per adapter. The values range from 1-8192 queue pairs.
|
Step 41 |
Server /chassis/adapter/host-eth-if/ #
set
rdma_resgrp
number
|
Sets the number of resource groups to be used. The values range from 1-128 resource groups.
Note
|
After committing the RoCE details, you are required to reboot the server for the changes to take place.
|
|
Step 42 |
Server
/chassis/adapter/host-eth-if #
scope
offload
|
Enters TCP
offload command mode.
|
Step 43 |
Server
/chassis/adapter/host-eth-if/offload #
set
tcp-segment-offload {disable
|
enable}
|
Enables or disables TCP Segmentation Offload as follows:
-
disable
—The CPU segments large TCP packets.
-
enable
—The CPU sends large TCP packets to the hardware to be segmented. This option may reduce CPU overhead and increase throughput
rate. This is the default.
Note
|
This option
is also known as Large Send Offload (LSO).
|
|
Step 44 |
Server
/chassis/adapter/host-eth-if/offload #
set
tcp-rx-checksum-offload {disable
|
enable}
|
Enables or disables TCP Receive Offload Checksum Validation as follows:
|
Step 45 |
Server
/chassis/adapter/host-eth-if/offload #
set
tcp-tx-checksum-offload {disable
|
enable}
|
Enables or disables TCP Transmit Offload Checksum Validation as follows:
|
Step 46 |
Server
/chassis/adapter/host-eth-if/offload #
set
tcp-large-receive-offload {disable
|
enable}
|
Enables or disables TCP Large Packet Receive Offload as follows:
-
disable
—The CPU processes all large packets.
-
enable
—The hardware reassembles all segmented packets before sending them to the CPU. This option may reduce CPU utilization and
increase inbound throughput. This is the default.
|
Step 47 |
Server
/chassis/adapter/host-eth-if/offload #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 48 |
Server
/chassis/adapter/host-eth-if #
scope
rss
|
Enters
Receive-side Scaling (RSS) command mode.
|
Step 49 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss
{disable
|
enable}
|
Enables or
disables RSS, which allows the efficient distribution of network receive
processing across multiple CPUs in multiprocessor systems. The default is
enable for the two default vNICs, and disable for user-created vNICs.
|
Step 50 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-ipv4 {disable
|
enable}
|
Enables or
disables IPv4 RSS. The default is enable.
|
Step 51 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-tcp-ipv4 {disable
|
enable}
|
Enables or
disables TCP/IPv4 RSS. The default is enable.
|
Step 52 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-ipv6 {disable
|
enable}
|
Enables or
disables IPv6 RSS. The default is enable.
|
Step 53 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-tcp-ipv6 {disable
|
enable}
|
Enables or
disables TCP/IPv6 RSS. The default is enable.
|
Step 54 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-ipv6-ex {disable
|
enable}
|
Enables or
disables IPv6 Extension RSS. The default is disable.
|
Step 55 |
Server
/chassis/adapter/host-eth-if/rss #
set
rss-hash-tcp-ipv6-ex {disable
|
enable}
|
Enables or
disables TCP/IPv6 Extension RSS. The default is disable.
|
Step 56 |
Server
/chassis/adapter/host-eth-if/rss #
exit
|
Exits to the
host Ethernet interface command mode.
|
Step 57 |
Server
/chassis/adapter/host-eth-if #
commit
|
Commits the
transaction to the system configuration.
Note
|
The changes will take effect upon the next server reboot.
|
|