Juniper BXOS 4.1 R2 - RELEASE NOTES 07-2010, BXOS 4.1 R2 Release Note

BXOS 4.1 R2 Software Release Notes

Release 4.1 R2
July 2010 These release notes accompany Release 4.1R2 of the BXOS software. They briefly

describe the hardware and software features, and identify the outstanding issues.
The BXOS software runs on the BX7000 Multi-Access Gateway.

Release 4.1 R2 Features

The following are a list of features for the BX7000 Multi-Access Gateway and the
BXOS Release 4.1R2. Following the description is the title of the manual to consult
for further information. For a complete list of manuals, see Table 3 on page 23.

Hardware  The gateway houses a Clock Card Expansion Bay that accommodates the

advanced clocking module.

 The advanced clocking module contains the following connectors for external

clock sources:

 2 BNC connectors

 1 RJ-45 connector

 For more information about the advanced clocking module, see the BX7000

Multi-Access Gateway Hardware Guide.

Interfaces  Support for IEEE 802.1ah OAM functionality on all Ethernet

interfaces—Ethernet interfaces support the IEEE 802.1ah standard for
Operation, Administration, and Management (OAM). The BXOS software
supports IEEE 802.3ah discovery and monitoring, remote fault detection, and
remote loopback. To enable IEEE 802.1ah support on the Ethernet interface,
include the

link-fault-management]

 Support for IEEE 802.1ag OAM functionality on all Ethernet

interfaces—Ethernet interfaces support the IEEE 802.1ag standard for OAM.
The IEEE 802.1ag specification provides Ethernet connectivity-fault
management (CFM), which is the standard for Layer 2 ping, Layer 2 traceroute,
and the end-to-end connectivity check of the Ethernet network. To enable IEEE

802.1ag support on the Ethernet interface, include the
the

interface statement at the [edit protocols oam-ethernet

hierarchy level.

interface statement at

[edit protocols oam-ethernet connectivity-fault-management] hierarchy level.

Release 4.1 R2 Features  1

BXOS 4.1 R2 Software Release Notes

Forwarding  Static or dynamic Ethernet pseudowire—Dynamic Ethernet pseudowire over

an MPLS or GRE tunnel is supported. Static Ethernet pseudowire over a static
MPLS tunnel is supported. An Ethernet pseudowire positioned between two
provider edges (PEs) carries the Ethernet frames or 802.3 protocol data units
(PDUs) from one PE to another over an MPLS network. The Ethernet
pseudowire operates in one of the three modes: raw mode, tagged mode, or
tagged-default mode.

 If an Ethernet pseudowire is operating in a raw mode, the

service-delimiting tags are never sent over the pseudowire. If a
service-delimiting tag is present when the frame is received from the
attachment circuit by the provider edge (PE), it is stripped (by the NSP)
from the frame before the frame is sent to the pseudowire.

 If an Ethernet pseudowire is operating in tagged mode, every frame sent

over the pseudowire must have a service-delimiting VLAN tag. If the frame
as received by the PE from the attachment circuit does not have a
service-delimiting VLAN tag, the PE must prepend the frame with a
dummy VLAN tag before sending the frame on the pseudowire. This is the
default operating mode.

For more information about the Ethernet pseudowire, see the BXOS Configuration
Guide.

 Increase in MTU size for Ethernet port interface (Routing mode)—The MTU is

configured for this interface by the user. In BXOS Release 4.1, the MTU value is
now limited to a maximum of 1900 bytes.

 Increase in MTU size for Ethernet port interface (Attachment Circuit (AC)

mode)— In BXOS Release 4.1, the user can configure the MTU value for this AC
port and the AC port will not receive or transmit more than the configured MTU
value. This MTU value is now limited to a maximum of 1800 bytes per Ethernet
frame.

 Increase in MTU size for Ethernet pseudowire—In BXOS Release 4.1, the MTU

value is increased to a maximum of 1800 bytes.

 Increase in number of concatenation of cells for ATM pseudowire— In BXOS

Release 4.1, the number of concatenation cells is increased from 28 to 34 cells
to utilize the available increased payload size with an increase in the
network-facing Ethernet interface MTU.

 Increase in MTU size for GRE tunnel interfaces— In BXOS Release 4.1, the

maximum MTU value is increased from 1500 bytes and is limited to 1900
bytes.

For more information about the change in MTU size, see the BXOS Configuration
Guide.

Timing and
Synchronization

 The BX7000 Multi-Access Gateway contains an advanced clocking module with

an internal oscillator that generates the clock signals that the BXOS software
uses. This clock provides Stratum 3 level accuracy and synchronizes the
time-division multiplexing (TDM) and Ethernet interfaces with various available
reference clocks.

2  Release 4.1 R2 Features

BXOS 4.1 R2 Software Release Notes

 The gateway supports the following internal input reference clock sources:

 T1 or E1 line clocks

 10 MHz

 BITS

 CES recovered clock

The gateway supports the following internal output reference clock sources:

 T1 or E1 line clocks

 Clock of 25 MHz for Ethernet interface

Routing and Signaling  BFD for static routes—The BFD for static routes is used to detect next-hop

failure in sub-seconds with less overhead. BFD can be used to alert any control
plane protocols about the failure. To enable BFD for static routes, use the

bfd-liveness-detection statement at the [edit static-route destination-prefix]

hierarchy level.

For more information about BFD for static routes, see the BXOS Configuration
Guide.

 VRRP implementation—VRRP implementation in BXOS provides the virtual

router function. All protocol messaging is performed using IP multicast
datagrams. A virtual router MAC address is used as the source of all periodic
VRRP messages sent by the master router to enable bridge learning in an
extended LAN.

A virtual router is defined by its virtual router identifier (VRID) and a set of
IP addresses. A VRRP router associates a virtual router with its real
addresses on an interface, and may also be configured with additional
virtual router mappings and priority for virtual routers it can backup. The
mapping between VRID and addresses must be coordinated among all
VRRP routers on a LAN.

Only the master for each virtual router sends periodic VRRP advertisement
messages. A backup router does not attempt to preempt the master unless
it has higher priority. VRRP router always becomes master of the any
virtual router associated with the addresses it owns. If the master is
unavailable, then the highest priority backup router transitions to master
after a short delay.

In BXOS 4.1R2 interface tracking feature is supported. Interface tracking in

VRRP group supports GIGE, VLAN, ML-PPP, PPP interface types and each
VRRP group can track a maximum of five interfaces. Bandwidth threshold
for interface tracking is not supported.

For more information about VRRP configuration including interface tracking
feature, see the BXOS Configuration Guide.

Release 4.1 R2 Features  3

BXOS 4.1 R2 Software Release Notes

Network and Device
Management

 Command-line interface—The BXOS command-line interface (CLI) is a

text-based interface that provides various commands for configuring and
monitoring the gateway. The CLI can be accessed through a console or network
connection using a secure shell (SSH) client. The CLI supports two modes of
operation: CLI operational mode and CLI configuration mode.

 SNMP support—The gateway supports SNMPv1, SNMPv2, and SNMPv3. SNMP

is supported for data integrity, confidentiality, and authentication. The gateway
accepts requests in any of the three SNMP formats.

 NETCONF—The gateway provides a NETCONF interface to enable retrieval and

manipulation of configuration data. It uses an Extensible Markup Language
(XML)-based data encoding for the configuration data and protocol messages.
NETCONF uses the XML interface for device manageability. NETCONF uses
SSH as the transport layer protocol and RADIUS or TACACS+ for
authenticating the SSH session. A maximum of four NETCONF sessions are
supported by the gateway.

 Zero touch configuration—The gateway downloads its configuration files from

a remote server and configures itself according to the configuration commands
in the configuration file. A USB disk drive can be used instead of a
configuration server to configure the gateway. If the zero touch bit is enabled
and a USB disk has the configuration file (

config.xml), the gateway uses this file

and does not make a Dynamic Host Configuration Protocol (DHCP) request.

 Notification through SNMP traps—The gateway sends events such as interface

up/down state, pseudowire up/down state, software upgrade status, and so on,
as SNMP traps to the configured trap destinations. The gateway supports both
SNMPv1 and SNMPv2 traps. The maximum number of trap destinations is
four.

Outstanding Issues

 Bootloader and OS image upgrade—The gateway supports upgrade of BXOS

and bxboot software through the CLI and the NETCONF interfaces. The
gateway supports FTP, TFTP, HTTP, or HTTPS protocols for downloading the
software image.

For more information about network and device management, see the BXOS
Configuration Guide.

 Unable to set the ARP timeout. The BX does not support ARP aging. There is no

workaround. [PR/273628]

 The VLAN subinterfaces in the gateway do not support the per-VLAN interface

statistics. The traffic forwarded through the VLAN interfaces is not displayed as
a part of statistics. There is no workaround. [PR/275851] [PR/303313]

 The gateway does not support moving a standalone PPP link into a multilink

PPP bundle. As a workaround, delete the standalone PPP link and reconfigure it
as a member link in the bundle. [PR/275939]

4  Outstanding Issues

BXOS 4.1 R2 Software Release Notes

 The gateway does not support moving a member link from one MLPPP bundle

to another bundle. As a workaround, delete the member link from the source
bundle and reconfigure the link as a new member link in the target bundle.
[PR/275941]

 You cannot remove the primary link you created on the gateway when there

are other links present in the bundle. As a workaround, remove all other links
created in the bundle before removing the primary link. [PR/275942]

 The gateway does not support configuring cell delay variance tolerance (CDVT)

for ATM VCs. There is no workaround. [PR/278451]

 BXOS does not permit Address Resolution Protocol (ARP) entries to age out,

but keeps them alive. Prior to aging out, the host operating system sends out
proactive ARP requests and keeps the entries alive. There is no workaround.
[PR/279807]

 The gateway does not have a provision to clear alarms generated by the power

supply unit (PSU). As a workaround, use the
the

[edit chassis] hierarchy level to clear the PSU related alarm. [PR/280393]

delete alarms power command at

 The BXOS software does not support OSPF virtual links on the gateway. There

is no workaround. [PR/281198]

 The gateway does not support tagged and untagged operations on a single

Ethernet interface simultaneously. To configure an Ethernet interface to use
tagged mode, you must create at least one subinterface before you associate
the Ethernet interface with an IP address. If you associate an Ethernet interface
with an IP address before you create a subinterface, the Ethernet interface
functions in untagged mode and you cannot create subinterfaces. As a
workaround, delete the IP address associated with the Ethernet interface and
create VLAN subinterfaces. This interface works only in tagged mode after this
operation. [PR/282223] [PR/282225]

 The BXOS software does not support IP fragmentation over PPP or MLPPP

interfaces. As a workaround, specify higher maximum transmit unit (MTU) size
of packets on the PPP interface so that the packets are not fragmented.
[PR/282469]

 The gateway does not support the octet-aligned T1 mode. There is no

workaround. [PR/283641]

 The gateway always sends IS-IS hellos packets padded to the MTU size. The

BXOS software does not support disabling IS-IS hello padding. As a
workaround, configure similar MTU sizes between IS-IS peers. [PR/285794]

 In BXOS, you cannot configure the IS-IS protocol as a level 1 or level 2 router.

By default, the IS-IS protocol is configured as both the level 1 and level 2 router.
As a workaround, disable level 2 in all IS-IS-enabled interfaces to configure the
gateway as the level 1 router or disable level 1 in all IS-IS-enabled interfaces to
configure the gateway as the level 2 router. [PR/285796]

Outstanding Issues  5

BXOS 4.1 R2 Software Release Notes

 The static MPLS tunnels assume that the next-hop through which the tunnel

destination is reached already has an ARP entry present in the ARP cache. If
you configure a static MPLS tunnel to a destination for which there is no
information in the ARP cache, the tunnel creation fails. As a workaround, use
static ARP entries for tunnel destinations before creating static tunnels.
[PR/286342]

 The gateway does not support jumbo frames. The maximum Ethernet MTU

support is 1500 bytes. There is no workaround. [PR/287113]

 The file delete * command deletes all files in the current working directory.

There is no workaround. [PR/287165]

 The BXOS CLI does not support a specific command to display the number of

PPP interfaces. As a workaround, use the

show interface command to display

the summary of all configured interfaces on the gateway. [PR/287229]

 The gateway does not support configuring interface alarms for a specific

interface. As a workaround, use the

set alarm command at the [edit chassis]

hierarchy level to configure alarms on T1, E1, or Ethernet interfaces.
[PR/287554]

 The VLAN priority bits can be marked only based on the VLAN subinterface

configuration. The gateway does not support the VLAN priority bits marking
based on the multifield (MF) or behavior aggregate (BA) classification. There is
no workaround. [PR/287562]

 The gateway does not support zero touch configuration from the JUNOScope

application. As a workaround, use CLI to configure the zero touch mode.
[PR/287964]

 The IP addresses for PPP interfaces are not dynamically negotiated with the

peer. As a workaround, configure IP addresses for PPP interfaces. [PR/287982]

 When you reboot the gateway after configuring and saving 200 pseudowires,

the gateway takes approximately 45 minutes to restore the configurations.
There is no workaround. [PR/289197]

 After configuring the LDP protocol, if you issue the show ldp overview

command, it does not display LDP configuration details. As a workaround, use
the

show protocol ldp command which displays LDP configuration details.

[PR/293862]

 The show route command takes approximately 5 seconds to display the output

if more than 100 routes are configured on the gateway. There is no
workaround. [PR/297833]

 In BXOS, path configuration is inside the LSPs. Therefore, you cannot use a

single path configuration for multiple LSPs. As a workaround, configure a path
for each LSP. [PR/301511]

 The global BFD protocol configuration that is optional is used for modifying the

default BFD parameters. Once you have configured the global BFD, you cannot
delete that configuration. As a workaround, set the global BFD parameters to
their default values. [PR/302179]

6  Outstanding Issues

BXOS 4.1 R2 Software Release Notes

 Pseudowire does not go down when the corresponding physical interface goes

down. There is no workaround. [PR/302247]

 In BXOS, VLAN subinterfaces do not support per-VLAN interface statistics. Any

traffic that is forwarded through VLAN interfaces is not displayed as part of the
statistics. As a workaround, use the

show interface command to display output

of traffic from a host. [PR/303313]

 When you create two RSVP LSPs, configure setup and hold priorities, and set

bandwidth ranges to these LSPs, the LSP with high setup and hold priority does
not preempt the other LSP. The available bandwidth value for each LSP is
reduced and advertised through routing protocols. When the bandwidth is
exhausted, new LSPs are not created. There is no workaround. [PR/305677]

 When traffic with packet size of 1500 bytes is sent at a speed of over 750 Mbps

or traffic with packet size of 64 bytes is sent at a speed of over 137 Mbps to the
gateway, the packets are dropped. The transmit-side packet scheduler does not
work when the incoming unidirectional traffic rate is very high. As a
workaround, keep the incoming packet rate below 750 Mbps for 1500
byte-traffic or below 137 Mbps for 64-byte traffic to view the effects of packet
scheduling. [PR/310273] [PR/453094] [PR/454145]

 If you configure a T1 or E1 interface, issue the show interface interface-name

command before sending traffic to that interface. The

rx_err_hec error counter

under traffic statistics increments and displays a non-zero value. As a
workaround, note down the initial header error control (HEC) error value after
configuring the interface and subtract the initial value from subsequent
statistics values to obtain an actual count of cells with HEC errors. [PR/311480]

 If a bypass tunnel is created for a peer loopback address, the backup tunnel

does not come up. If the bypass tunnel is created for a peer loopback address
where a primary tunnel also ends, then the backup tunnel is not created over
that bypass tunnel. As a workaround, configure the bypass tunnel to the peer's
physical interface address (instead of the loopback address). [PR/312644]

 The BXOS software does not support the show statistics command for

displaying the configured scheduler. There is no workaround. [PR/312830]

 The CLI does not display statistics of packets that are shaped using traffic

shapers. There is no workaround. [PR/312831]

 jnxPwEnetVlanMode table does not work for Ethernet PWE through

GigabitEthernet backhaul. The DUT does not display the proper VLAN mode
when the Ethernet PWE is configured with the egress options of None, Strip,
and Replace. Instead of changeVlan, removeVlan is displayed. As a
workaround, IM-DPM structures are used instead of DCL. [PR/314788]

 In the absence of a time-zone command, the system time should be set to the

local time. There is no workaround. [PR/314980]

 Dynamic changes to a loopback address are not advertised to the peer. As a

workaround, disable OSPF, change the loopback IP address, and then enable
OSPF. [PR/387049]

Outstanding Issues  7

BXOS 4.1 R2 Software Release Notes

 When more than 5 IS-IS sessions are configured on VLAN interfaces with a

default hello-interval of 3 seconds and hello-multiplier of 10 seconds; ISIS
sessions will flap and cause the BX7000 to crash. As a workaround, configure
the ISIS hello-interval to 15 seconds and the hello-multiplier to 5 seconds.
[PR/387294]

 When you configure OAM on a switched connection and send a continuous

stream of OAM cells to the gateway, the OAM cells are not processed but are
dropped. There is no workaround. [PR/389071]

 After you configure the link protection and pseudowires over the protected

tunnel, if the primary link is disabled on the gateway side, the traffic switch
times are longer than expected. There is no workaround. [PR/389024]

 When you configure fast reroute (FRR), the traffic failover time from the

primary tunnel to the backup tunnel is approximately 300 milliseconds. There
is no workaround. [PR/389024] [PR/410625]

 During the BXOS image upgrade, when you issue the request system reboot

command from the serial console of the gateway, the image upgrade process
stops. As a workaround, do not reboot the gateway during the image upgrade.
[PR/390219]

 GRE interface configuration requires the input of tunnel parameters along with

interface creation. You cannot create the interface without any related
parameters, and then associate the tunnel parameters at a later point in time.
Similarly, after being created, the parameters cannot be modified. As a
workaround, delete and recreate the interface. [PR/392812]

 When you create or delete GRE or RSVP tunnel, the operating system does not

free the allocated memory completely. There is no workaround. [PR/392833]
[PR/392835]

 The DUT crashed and rebooted with error messages while executing show ospf

interface detail command with 210 OSPF with MD5 configurations over 210

VLAN interfaces (distributed across 7 Gigabit Ethernet ports including MB and
Addon with 30 units per interface). There is no workaround. [PR/393750 ]

 The show mpls tunnel <tunnelname> association command displays the status

of the tunnel as up even if the tunnel is down. The system tries multiple times
to find the backup tunnel for the protected tunnel. If the system does not find
the backup tunnel, after few minutes, it will declare the tunnel as down. There
is no workaround. [PR/397034]

 The IPv4 fragmentation feature does not work. As a result, packets that are

greater than the MTU of the outgoing Ethernet interface are dropped. There is
no workaround. [PR/400112]

 The traffic with packet size greater than the configured MTU value can pass

over a GRE interface. As a workaround, AAL5 SDU/PDU PWEs can be created
with the PWE MTU to be 12 bytes less than that of the backhaul interface MTU
for PW over RSVP tunnel, and PWE MTU must be less than the GRE tunnel
MTU for PW over GRE tunnel. [PR/400194]

8  Outstanding Issues