OS6450 AOS 6.6.4.309 R01 Release Notes
OS6450 AOS 6.6.4.309 R01 Release Notes
OS6450 AOS 6.6.4.309 R01 Release Notes
R01 - 6/2/2015
Release 6.6.4.309.R01
The following is a list of issues that have been identified and corrected in AOS
software release. This document is intended to be used as a pre-upgrade
guide and does not replace the Release Notes which are created for every GA
release of software.
Important Notice: For a copy of software release not posted on the Web or if you have any question
or concern please contact Alcatel’s Technical Support Department.
1 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
2 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
3 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
4 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
5 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
6 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
7 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
9 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
10 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Under Verification:
PR 166633 Build: 6.6.4.37.R01
Summary: Missing link after polling switch in Omnivista 3.5.2 version.
Explanation: Prevent mismatch of chassis ID on CMM and NI after takeover
11 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
12 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Explanation: Cosmetic issue fixed such that correct values are updated while extracting any of the
power supply from the unit
14 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
15 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
16 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
18 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Explanation: The check for the bounds of lacp agg actor port priority has been changed to 0 to
255.
19 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
20 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
switch
Explanation: As per our analysis the root cause of the issue is whenever aaa send command
message to server for processing the accounting request, the aaa command
accounting will use the maximum size of command length which is 512.but when aaa
command accounting is configured as local, it is using the buffer of size 255 because
of this local accounting server is not able to hold the entire values of accounting
command which also makes the switch to crash.so changes have been made to
increase the buffer size as same as accounting command
22 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
OS6450 Stack issue: Primary CMM MAC is not getting synchronized correctly to
Summary:
IDLE Unit.
Explanation: IDLE unit Interface MAC is correctly synced with the Primary CMM MAC in the
hardware table.
23 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
24 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Known Issues:
PR 198904
Summary: OS6450 is not forwarding the OSPF hello packet to other port with Ethernet-service.
Explanation: Enable/disable the ip multicast routing status in hardware on vlan basis. By default it
should be disabled for all vlans.
Introduction:
CPE test head feature shall now support bi-directional traffic functionality. This bidirectional functionality is
achieved via unblocking the loopback mode in the CPE test head feature. In this mode Generator and
Analyzer DUT will be same and remote DUT will configure as a Loopback DUT. The remote device shall
reflect the traffic back to the originating device. The originating devILice shall count and drop the reflected
traffic. This shall help measuring the network performance of customer traffic across network at a single end
point. This feature shall support up to eight concurrent streams configured under test groups.
The CPE Test Head feature include one way test with the ability to report the results from the remote device
on the unit initiating the test. This would also automate the way in which the test is triggered on the remote
device. This feature shall add option to trigger the start at the remote end from the sender side via a
proprietary protocol. Also the generator device for the test shall have a mechanism to gather Rx-Ingress test
counters from the remote device and store it in a local database at the end of the test.
The CPE Test Head shall measure RTT and jitter during the test head operation. The L2 SAA test shall run
between two supporting Omni switches. The L2 SAA tests shall run alongside the data traffic tests. The test
results shall be available at the initiating device. This feature shall be available to both unidirectional and
bidirectional tests.
Platforms Supported:
Omni Switch 6450
Omni Switch 6250
25 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Commands usage:
Syntax Definitions
Role The DUT shall have one role configured for a test-oam group as
Generator, analyzer or loopback.
test-oam <string> { [vlan <vid>] [port <slot/port>] [packet-size <size of packet in bytes>] start|stop} [fetch-
remote-stats]
While the fetch-remote-stats option is used, the test at the remote end shall be triggered from the generator
side, stats shall be collected at the end of the test and finally test shall be stopped after receiving the test
results.
Syntax Definitions
String The string is an identifier of the traffic test. Up to 32 tests can be configured. The string
Can be of length 1to 32 characters long.
<slot/port> The port to be used in the testing. Depending on the role this port shall have different
Interpretation. Generator -> port generating the frame. Loopback -> this port shall be the
Port where Loopback of the traffic shall take place. Analyzer -> Port configuration is not
Required.
<Size of packets in bytes> the size of packets in bytes, it can be of size from 64 byte to 9212
Bytes. It shall also include the size of CRC. Default value is 64.
Fetch-remote-stats When this option provided, remote start/stat feature is enabled. User shall be
Able to start the test at the remote end from the generator side and also shall be
Able to collect the RxIngress counter results from remote at the end of the test.
test-oam <string> L2-SAA [priority <vlan-priority>] [count <num-pkts>] [interval <inter-pkt-delay>] [size <size>]
[drop-eligible {true|false}]
While the cli is configured, the Testoam shall run SAA tests in parallel with the test streams.
Syntax Definitions
26 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
String The string is an identifier of the traffic test. Up to 32 tests can be configured. The string
can be of length 1to 32 characters long
vlan-priority This is to specify both the internal priority of the Mac ping and the 802.1p value on the
vlan tag header. Default is 0
count The number of packets to send in one ping iteration. Default value is 5.
inter-pkt-delay Delay between packets sent during a ping iteration in milliseconds. Default value is 1000ms.
size The size of the payload to be used for the MAC- ping iteration. Default value is 36 bytes
drop-eligible This is to specify both the internal drop precedence of the MAC ping and the CFI bit on
the vlan tag header. Default is false.
Usage Guidelines
L2-SAA test shall derive the source mac, destination mac and the vlan id from the testoam configuration for
individual test streams. The user shall be able configure a different SAA profile for each individual stream.
Default L2 SAA configs shall be applied when no optional parameters are provided.
Syntax Definitions
enable This will enable the logging of test-oam statistics to the file.
disable This will disable the logging of test-oam statistics to the file.
Limitations:
When Test stream rate is configured as line rate, the remote statistics (the traffic statistics which gets
reflected back from remote DUT configured as LOOPBACK) will differ from the traffic statistics generated
from the DUT configured as GENERATOR.
Introduction:
This feature Enhancement provides the facility to increase the buffer size or change the profile for reducing
the drops in the traffic. The OS6250/6450 comes with pre-canned buffer settings for all ports. This means that
each queue on each port is statically set with a guaranteed number of buffers and descriptors.
In order to accommodate intermittent bursts, the switch supports a shared pool of buffers and descriptors that
allows a queue to use a shared resource when its guarantee resource is exceeded. Neither the queue
guaranteed resources or the shared resources are configurable. This implementation gives a limited buffering
capability to support burst of traffic without discarding traffic. For some customers who are using specific
applications, this is a major limitation.
27 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Taking this into consideration, this enhancement attempts to overcome the limitation by increasing the
buffering capacity, i.e. the ability to modify the number of buffers in the shared pool. The enhancement also
adds the ability to change the default buffer profile assigned to the ports.
Platforms Supported:
Omni Switch 6450
Omni Switch 6250
Commands usage:
Syntax Definitions
Integer The Integer specifies the number of shared buffers
Range {0 – 4095}
Default 1500 [applied since switch boot up]
28 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Limitations:
Hardware Limitations
The buffer settings are limited and do not expose the entire ASIC configuration
The number of Tail Drop profile is not exposed. The content of the internal profiles is not displayed nor can
the internal profiles be modified
The Tail Drop settings (enable/disable, random tail drop) are not exposed and stay to their default value
The Global System Limits (buffer, descriptor) are not exposed and stay to their default value
The Resource Sharing settings (resource sharing for DP1, resource sharing for individual queue) are not
exposed and stay to their default value
The Per Port Limits (buffer limit, descriptor limit) are not exposed and stay to their default value
Software Limitations
If the Port Profile assignment fails in hardware, an error is displayed on the console and also reported in QOS
logs. There may be a mismatch in port profile assignment displayed in the “show qos register” command
output and the value applied in hardware.
29 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Introduction:
This Feature is implemented to increase the ARP entries limit to 1K(1024).During the boot up based upon the
role of the devices Metro / Non Metro the ARP limit is decided. If it is a non-Metro device the number of ARPs
is limited to 1024 entries.
1K ARP support based on the model type and installed license during boot-up
- For OS6250-M model, no changes and ARP is limited to 256
- For OS6250/6450 without a metro license, support 1K ARP
- For OS6250/6450 with metro license, no changes and ARP is limited to 256
Earlier, the number of ARP for Metro as well as Non-Metro was limited to 256.
Platforms Supported:
Limitations:
512 ARPs would be supported in Hardware, while remaining 512 ARPs would be added only in software and
not in Hardware. Traffic for ARPs resolved in software would only be routed in software, so these traffics
would cause CPU spike and packet loss.
Introduction:
Prior to this enhancement command authorization in TACACS is done based on partition-management family
that the command belongs to. According to the new feature, after authentication, once command based
authorization is enabled then every cli command that the user executes on the switch is sent to the TACACS+
server. So TACACS+ server will do the authorization for the whole command and send the RESPONSE
message to the TACACS+ client. If command based authorization is disabled then PM family for the
command is sent for the authorization.
Platforms Supported:
Omni Switch 6250, 6450
Commands usage:
aaa tacacs command-authorization {enable/disable}
By default command authorization is disabled
Configuration snapshot:
1. aaa tacacs command-authorization disable
30 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Limitations: Snmp and http are not supported in Command based authorization
Introduction
This feature Enhancement provides the facility to enable/disable the console cli session so that access to the
switch configuration shell through the console port can be in a controlled manner as required. By default this
facility (console access) will be enabled. This can also be stored in configuration file so that console access
control can be applied even after reboot.
31 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Recovery Procedure
If both the console cli session is disabled in the configuration file on both working and certified directory and if
the telnet/ssh/web view session is not available to the switch, to get access to the switch console cli session
user have to stop the switch in miniboot by setting the boot flags to 0x1000 and once the switch stops in
miniboot user shall delete the configuration file and reboot the switch to get console access to the switch.
Earlier, there was no provision to control the access for console cli session.
Platforms Supported
Omni Switch 6250, 6450
Usage Guidelines:
By default, the cli console shell is enabled. The command shall be accepted only via Telnet/SSH session, and
not through console sessions to the switch. When it is disabled, even the switch log output to the console
shall be disabled. Command shall be stored to the configuration file using write memory. Command shall be
used only on standalone unit, even if used in stack only primary unit console CLI session shall stay disabled.
Limitations:
None
Introduction:
This feature Enhancement provides the ‘Standard VLAN support on NNI ports’ will allow any standard (non-
service) VLAN to be associated to NNI ports. This allowed association can be of type untagged or 8021q
tagged. However, there is an exception for VLAN 1, which shall not be associated as untagged member to a
NNI port. This will allow the customers to configure 802.1q services, QinQ service and untagged services
using the same uplink NNI port. This will also allow the customer to use an untagged management VLAN to
manage the switch via NNI ports.
With the implementation of this feature, the following will be the changes on the behavior of the switch:
The standard VLAN configuration (both untagged and 802.1q tagged association) will now be allowed on an
NNI interface binded with a service VLAN.
The binding of service VLAN to NNI interface will now be allowed when the interface (physical or linkagg) is
already tagged with standard VLAN.
802.1q VLAN tagging to an NNI interface will not be allowed if the interface is set with TPID other than
0x8100.
Any modification with respect to TPID will not be allowed if the NNI interface is 802.1q tagged.
There would also be significant changes with respect to the default VLAN of the NNI interface (both physical
and LAG):
If an interface is already an untagged member of a standard VLAN other than VLAN 1, then on making it an
NNI interface, there should not be any change with respect to the default VLAN of the interface. (Currently,
the default VLAN changes to 4095).
32 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
If the default VLAN is removed from the NNI interface, then the default VLAN should be changed to 4095
It implies, from the above two points, that it shall not be possible to configure VLAN 1 as default VLAN of an
NNI interface.
Platforms Supported:
Omni Switch 6450
Omni Switch 6250
Commands usage:
33 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Limitations:
None
Introduction:
This enhancement has increased the number increased of telnet sessions from 4 to 6, no of syslog servers
increased from 3 to 12 and no of NTP servers increased from 3 to 12
Platforms Supported:
Omni Switch 6450, 6250 (Non-Metro)
Commands usage:
No new CLI introduced for this.
Limitations:
None
Introduction:
This feature enhancement facilitates to support the link layer classification in order to interoperate with newer
class 4 PD’s(Powered Device) , because these devices require a response to the LLDPDU power via MDI
TLV before they will draw additional power from PSE(Power). Earlier the maximum power is set to the
maximum allowed power for the detected Power Class of the Power Device connected on the port. The
Power Class detection is done via hardware by the POE controller. POE Devices in general can draw any
amount of power up to the maximum power that is set for the port. In any condition, the maximum power that
the PD can request from the PSE cannot exceed the maximum allowed power for the Power Class in which
the PD(Powered Device) is detected, But these newer class 4PD’s(Powered devices) requires to draw
additional power than the maximum power set for the port Hence this feature is introduced.
34 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Platforms Supported:
Omni Switch 6250, 6450
Commands usage:
lldp {slot/port | slot | chassis} tlv dot3 power-via-mdi {enable | disable}
Syntax Definitions:
slot/port Slot number for the module and physical port number on that module
slot The slot number for a specific module.
enable Enables 802.3 TLV LLDPDU transmission.
disable Disables 802.3 TLV LLDPDU transmission.
Usage Guidelines:
• The LLDPDU must be enabled and set to transmit before using this command.
• If this command is applied to a slot or chassis, then the existing configuration related to this command
is lost.
Limitations: None
Introduction:
This feature Enhancement provides the facility to Establish a SSH Session for a Read-Only Users through
Switch as Local Server, Radius Server,LDAP,TACACS.This SSH Read-Only Session allows to view the SSH
Specific show commands .
Earlier it is not possible to SSH to a switch and access for a user unless he has read-write permissions.
This is the current default behavior. But telnet to switch does not validate the permissions of the user and
therefore switch becomes accessible.
Platforms Supported:
Omni Switch 6450, 6250
Commands usage:
user {username} read-only ssh password {maximum 8 }
Syntax Definitions:
Read-only :Specify the User Privilege
Ssh: The type of Service and Family the Belong to
Defaults
Parameter Default
Read only for families None
Usage Guidelines
Read-Only user configuration must specify the SSH family
Creating a user with Family as “none “ will not permit access to SSH
35 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Show users:
Displays information about the all the user configuration
Alcatel-Lucent ESD – IP Networking – Service Release – 6.6.3.509.R01 - 9/11/2013
Examples:
User name = goog,
Password expiration = None,
Password allow to be modified date = None,
Account lockout = None,
Password bad attempts = 0,
Read Only for domains = ,
Read only for families = ssh ,
Read/Write for domains = None,
Snmp allowed = YES,
Snmp authentication = NONE,
Snmp encryption = NONE
Limitations: None
Introduction:
This enhancement provides facility to detect a stack split via the device acting as helper, called as SSP
Helper (Stack Split protection Helper).For the device to act as helper we need to explicitly enable the helper
mode, and it should be connected to the stack via linkagg.
The basic functionality of the helper would be to transmit the received health PDU to other ports in the linkagg
associated with SSP. When enabled all the linkagg ports associated with the SSP would be programmed for
receiving the SSP PDUs. When the SSP PDU is received the Helper NI would send SSP PDU to all other
ports in Linkagg. When Helper receives SSP PDU with protection mode, it would immediately send an
acknowledge of Protection mode receive and forward the protection mode PDU to all SSP ports.
Platforms Supported:
Omni Switch 6450
Commands usage:
Description: Linkagg-id on which to apply the SSP protocol on linkagg member ports for helper device.
Description: This command shows SSP Helper status of the Link Aggregation ID assigned.
Configuration Snapshot:
1. Show stack split-protection helper enable/disable:
Limitations:
None
Introduction:
Prior to this enhancement the mac tunneling feature can be enabled/disabled by setting the variable in
AlcatelDebug.cfg. The functionality remains same but introduced the cli to enable and disable the mac
tunneling feature. If the mac tunneling is enabled the destination mac in the frame is replaced with tunnel
mac. If status is disabled there will not be any change in destination mac of the Frame.
37 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Platforms Supported:
Omni Switch 6450
Omni Switch 6250
Commands Usage:
Usage guidelines
By default, mac-tunneling is enabled.
The command will take effect only after write memory and reload of the switch.
While changing the status the below info message will be displayed. (INFO :Changed mac-tunnel
feature status will take effect if command is saved on next switch reboot)
Limitations:
The command will take effect only after write memory and reload.
12. DHCP snooping binding table for IP source filtering enabled ports
Introduction:
Prior to this, command will display the ports or vlans on which the ip source filtering is enabled. Added
additional option “binding” to display the binding table for ip source filtering enabled ports. The binding table
output is same as the output of show ip helper dhcp-snooping binding but it will show binding table for ip
source –filtering enable ports.
38 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Platforms Supported:
Omni Switch 6450
Omni Switch 6450
Command Usage:
show ip helper dhcp-snooping ip-source-filter {vlan| port |binding}
Syntax Definitions:
vlan Displays the VLANs on which the IP source filtering is enabled port
Displays the ports on which the IP source filtering is enabled binding
Displays the binding table for the ports on which the IP source filtering is enabled
Usage Guidelines:
The show output displays only those ports or VLANs on which IP source filtering is enabled
or binding table for ip source filtering enabled ports.
This command also displays the status of the link aggregate ports when source filtering is
enabled at VLAN or port level.
Snapshot:
Limitations:
None
Introduction:
This feature enhancement facilitates to configure policy rule that specifies rate limiting as action for a group of
ports or individual ports as per our requirement. For this enhancement new attribute “split & non-split” has
been added for a policy port group to specify whether the group needs to be treated as a list of individual port
or not respectively. This feature provides the following two modes to be applied as a part of the policy source
port group:
1. Non-split: When used with this mode, the rule for rate limiting is applied for the group of ports. This is the
default behavior for the source port group.
2. Split: When used with this mode, the rule for rate limiting is actually applied for each of the individual ports.
39 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
However, the action is not restricted to rate limit the incoming traffic, action could be anything other than the
keyword “share”. Moreover, other actions can also be applied in addition to rate limiting, such as changing the
dscp value, etc. Any incoming traffic in access of the applied bandwidth to an individual port will be dropped.
Before this enhancement, on configuring a policy rule that specifies a rate limiter as action and a source port
group as condition, the rate limiter is actually applied for the group of ports, not each individual port.
Platforms Supported
Commands usage
Syntax Definitions
split In this mode, the rule for rate limiting is actually applied for each of the individual ports.
non-split In this mode, the rule for rate limiting is applied for the group of ports. This is the default behavior
for the source port group.
Usage Guidelines
When the port group is configured in the split mode, the rule needs to be split into multiple sub-rules.
Depending on the policy condition for the rule, each sub-rule may consist of multiple entries
The rate limiter is to be shared between the entries for the same sub-rule.
40 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Limitations:
The scope of this feature is limited to source port group can be attached to only default policy list. Any rule
with the source port group in the split mode attached to policy list will throw an error.
Introduction:
This feature Enhancement provides the ‘Tri speed SFP support on OS6450 U24’.The Copper Small Form
Pluggable(SFP)s Finisar FCLF 8521-3 and Finisar FCLF 8521 P2BTL are compatible with Gigabit
Ethernet(1000 Mbps), Fast Ethernet (100 Mbps) and Ethernet(10 Mbps).
Platforms Supported:
Omni Switch 6450
Commands usage:
show interface: show configuration snapshot interface
Syntax Definitions
Configuration snapshot interface verify the configuration of the interface
Limitations:
None
41 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Introduction:
The configuration file management feature is to modify the configuration file label corresponding to the
directory it resides, without affecting any functionality. Earlier when configuration file is retrieved from working
and certified directories of Omni switch, they all have the same label as in old directory in the beginning of file
regardless if you retrieve the file in working or certified directory. So after retrieving, it’s difficult to find from
where the configuration file belongs.
The operations of existing configuration file management system:
While performing certify and/or synchronization or restoration process in Omni switch the configuration file of
source directory will be copied to the destination directory based on the below conditions.
a) If the configuration file doesn't exist in the destination directory.
b) The file exists but differs in size and/or time stamp.
If any of the above condition is true, the configuration file will be copied to the destination directory and the
timestamp of source directory configuration file will be re-applied on the copied configuration file in destination
directory.
After the source configuration file contents copied to destination configuration file, the label in destination
configuration file will be modified and the time stamps of source configuration file will be re applied.
Platforms Supported:
Commands usage:
While executing the commands in the below table configuration file header should be updated showing the
directory it is located and re-apply the source directory configuration file timestamp.
42 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Expected Outcome:
After issuing certify/synchronization commands mentioned in table 1, the process should complete without
any errors and the label inside the boot.cfg file of certified directory should contain certified directory in the
label.
Sample output:
After issuing commands for restoring the files (mentioned in table 1), the process should complete without any
errors and the label inside the boot.cfg file of working directory should be remain unchanged.
Sample output:
There should not be any functional impact on existing synchronization status determining logic.
Sample output:
Limitations: None
Introduction:
Remote Fault propagation (RFP) propagates connectivity fault events into the interface that is attached to a
MEP. Once the fault is detected for a MEP, the MEP’s interface is shutdown. Unlike other violation
mechanisms that keep the link up when an interface is shutdown, this fault propagation mechanism will
effectively shutdown the link so that the remote end of the interface also detects a link down.The feature is
43 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
configurable on per MEP basis and is supported only for UP MEPs. Remote Fault Propagation detects only
Loss of connectivity and Remote MAC defect.
Platforms Supported:
Omni Switch 6450
Omni Switch 6250
Commands usage:
ethoam endpoint <mep-id> domain <md-name> association <ma-name> rfp {enable|disable}
Above CLI shall enable or disable RFP on MEP
Syntax Definitions
<mepid> A small integer, unique over a given Maintenance Association, identifying a specific
Maintenance association End Point. MEP-ID is an integer in the range 1-8191.
<md-name> Domain name.
<ma-name> Association name.
Usage Guidelines
The domain and association must be created before RFP can be enabled.
The end point must be configured in the MEP list, before it can actually be created.
The MEP must be an UP MEP. If down MEP is specified, CLI returns with an error.
The admin state of the MEP must be enabled in order to report faults.
RFP cannot be enabled on virtual UP MEP since it is not associated with a physical interface.
If RFP is enabled on an UP MEP created on a linkagg, then detection of RFP violation will shutdown the
individual member ports. No new ports should be added to or removed from the linkagg at this time. This will
not be blocked from configuration, but is left to the user.
It is recommended that if RFP is enabled on a port, then any other violation feature (Link Mon or LFP) should
not be configured.
It is recommended that if RFP is enabled on a port, then automatic recovery is disabled for that port.
If Link Mon is configured on a RFP enabled port, then the WTR timer must be less than the CCM interval.
Example:
Syntax Definitions
<mepid> A small integer, unique over a given Maintenance Association, identifying a specific
Maintenance association End Point. MEP-ID is an integer in the range 1-8191.
<md-name> Domain name.
<ma-name> Association name.
44 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Example:
Limitations: None
Introduction
Currently the SSH key size 512. The SSH key size for certificate generation will be increased from 512 to
1024 for additional security. The switch uses default certificate for establishing its identity when acting as web
server (receiving http/https requests). This certificate is generated once and stored in the flash as wv-
cert.pem.
Since the certificate is stored persistently in flash, to allow the new key size to take effect the certificate needs
to be regenerated. This means the certificate file needs to be deleted and switch rebooted in order to use this
feature post upgrade to this release. The new certificate will be generated with increased key-size
Limitations
This key size increase is only applicable to AOS auto generated certificate. Customers using their own
certificates need to ensure that they are generated with proper key size.
45 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Platforms: OS6250,OS6450
In AOS, IPv4 and IPv6 multicast protocols are by default always copied to CPU. The high CPU usually
impacts the normal operations of the Omni Switch protocols such as LACP, ERP.
In Order to resolve this high CPU issue, this feature is introduced to control the processing of the IPv4
multicast protocols.
The processing of all IPv6 multicast protocols is globally controlled by the presence of an IPv6 Interface.
No IPv6 interface configured
All protocols in the ff02:0::/32 range are transparently forwarded and not copied to CPU.
At least one IPv6 interface configured
All protocol packets in the ff02:0::/32 range are copied to CPU on all vlans irrespective on which vlan
IPV6 interface is enabled.
MLD packets are copied to CPU based on the global ipms status. When IPMS is globally enabled, MLD
packets are copied to CPU. When IPMS is globally disabled, MLD packets are not copied to CPU.
Command Usage:
1. To enable/disable multicast dynamic-control drop-all status
ip multicast dynamic-control drop-all status [{enable|disable}]
Guidelines: By default this status is disabled. If it is enabled, all ipv4 multicast packets including ipv4
multicast well-known protocol packets will be dropped. IPv4/IPv6 multicast protocol packets are given below
in Note section.
Note:
If this command is entered without any enable/disable option, disable action will be applied.
Below are the well-known IPv4/IPv6 multicast protocol packets,
VRRP: 224.0.0.18/32 + IP protocol 112
RIPv2: 224.0.0.9 + UDP port 520
Examples
->show ip multicast
Status = enabled,
Querying = enabled,
Proxying = disabled,
Spoofing = disabled,
Zapping = disabled,
Querier Forwarding = disabled,
Flood Unknown = disabled,
Dynamic control status = disabled,
Dynamic control drop-all status = disabled,
Buffer Packet = disabled,
Version = 2,
Robustness = 7,
Query Interval (seconds) = 125,
Query Response Interval (tenths of seconds) = 100,
Last Member Query Interval (tenths of seconds) = 10,
Unsolicited Report Interval (seconds) = 1,
Router Timeout (seconds) = 90,
Source Timeout (seconds) = 30,
Max-group = 0,
Max-group action = none
Helper-address = 0.0.0.0
Limitations
The proposed solution does not address the DOS attack concern
Injecting a high rate of well-known protocol on a port will still cause a high CPU.
Dynamic-Control “drop-all” feature should not be enabled if a routing protocol or VRRP is configured
on the Omni-Switch as protocol packet will be dropped.
The basic idea of this feature is to convert the untagged frames into double tagged frames in the provider
network so as to make ICMP between the endpoints to work. The frames should be always untagged on the
customer network. This will be ensured using double push and double pop operations. The double push will
happen on the UNI port in order to push the configured CVLAN as well as the SVLAN in the egressing packet.
The double pop must be applied on the NNI port in order to remove both the tags when the packet is
egressed from the UNI
47 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Usage
Examples
ethernet-service untagged-cvlan-insert enable
ethernet-service svlan 10
ethernet-service svlan 10 nni linkagg 1
ip interface test address 10.10.10.2/31 vlan 10 cvlan 20
ip interface "Loopback0" address 10.10.10.5
Limitations
Enabling “Cvlan insertion for untagged packets” feature on the switch would imply that the existing
legacy behavior of UNI and NNI ports will no longer hold good
Control traffic other than IP traffic destined to the switch out of scope of this feature
The “show ip interface” will not display the mapped interfaces.
The feature is meant for all IP traffic which is supported by the switch. Any other traffic which in-turn
goes through the same interface will also be double tagged.
As CVLAN-SVLAN is a one to one mapping, only one interface which uses the same SVLAN can
hold the CVLAN. When we try to create another interface using the same SVLAN, and try to give a
CVLAN value, it is expected to throw an error.
CVLAN tag is supported only for normal interfaces and not for dhcp-client ip addresses
48 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
Platforms: OS6450,OS6250
In the existing implementation, AOS uses the default SSH TCP port (port 22) to establish an SSH session.
With the new implementation, when the user configures the TCP port number for SSH session, it will be
saved in the switch file “/flash/network/sshConfig.cfg”. In order to use the configured port number while
establishing the SSH session, the switch must be rebooted.
While the switch boots up, if the file “/flash/network/sshConfig.cfg” exists, it will be parsed to read the TCP
port number that should be used to establish the SSH session, otherwise the default SSH TCP port shall be
used.
Usage
Command to configure TCP-PORT number for establishing SSH Session.
Note: Well-known reserved TCP port numbers and the IP ports which are internally used in AOS are excluded
in assigning to SSH TCP port.
Limitations
Switch must be rebooted after configuring the TCP port number so as to use the configured TCP port
number when establishing SSH sessions.
Well-known reserved TCP port numbers(ports 20,21,23,25,69,80,161,389,443) and the IP ports which
are internally used(defined in system_ipport.sh) are excluded in assigning to SSH TCP port. Error will
be thrown when these ports are tried to be configured for SSH port.
21. TWAMP
Platforms: OS6450,OS6250
Two-Way Active Measurement Protocol (TWAMP) provides a standard technique to measure network
performance metrics. Unlike ICMP Ping, TWAMP also measures round trip delay/Jitter apart from the RTT.
Moreover TWAMP does not require clock synchronization between the two devices. The initial release will
49 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
support the TWAMP Server and/or Reflector Implementations of TWAMP in Unauthenticated Mode only for
IPv4.
Following are the functionality provided by the feature.
TWAMP server would transmit a test packet to the Session-Sender in response to every received
packet
AOS S/w also implements a REFWAIT timer functionality to monitor inactivity in test sessions.
loopback0 IP address configured on the switch will be taken as the IP address of the TWAMP Server.
Usage
1. Respond to TCP Open messages from various clients and establish TWAMP ControlConnection
DUT
Limitations
1) Time-stamping is not available in hardware on all platforms. Hence time-stamping is done in software on all
platforms, namely Kite-2, Etna, Stackable Etna, Fuji, Fuji-2 and Garuda.
2) The TWAMP operations will use software based timestamps and hence will not provide precise
measurement of network delay.
3) The TWAMP Server/ reflector will not use the DSCP of the Control- Client’s TCP SYN in ALL subsequent
packets on that connection (control and test packets).
4) The statistics displayed in “show twamp server connections” command is updated on a regular time interval
only
Network Address Translation (NAT) is a feature that allows an organization’s IP network to appear from the
outside to use different IP address space than what it is actually using. Thus, NAT allows an organization
51 / 54
Alcatel-Lucent Enterprise – IP Networking – Service Release – 6.6.4.309.R01 - 6/2/2015
which uses private addresses (local addresses), and therefore not accessible through the Internet routing
tables, to connect to the Internet by translating those addresses into globally routable address space (public
addresses) which are accessible from Internet. NAT also allows organizations to launch readdressing
strategies where the changes in the local IP networks are minimum. NAT is also described in RFC 1631
Network Address Translation (NAT) is used for rewriting a source or destination IP address to another
address. A single address may be rewritten, or an entire subnet or list of IP addresses may be rewritten to a
group of addresses.
1) Static NAT is where the mapping of local and global addresses is unanimous.
2) Dynamic NAT is a mapping of local addresses in a pool of global addresses. This means that the
mapping between global addresses and local addresses is not unanimous and depends of the
execution conditions.
3) NAPT (Address Port Translation) is mapping between local addresses and a unique global address.
In this case a translation of the transport protocols ports (UDP, TCP) is carried out.
Usage
The source/destination ip/network should be an interface ip on the NAT device which needs to be
NAT’ed.
Example
1. The policy nat will rewrite the source address for any traffic from the 10.0.0.0 network to the
Internet friendly address, 143.209.92.42
2. Traffic destined for the 10.0.0.0 network will be rewritten to the original IP addresses based on
the dynamic TCP/UDP port assignment
NAT DUT1:
Limitations
54 / 54