• Tag : 4.21.3F

 
 

BGP neighbor default-originate always

Description Neighbor default-originate feature is used to advertise a default route to the neighbor ( peer or peer-group ) even when a default route is not present in the BGP RIB. If a default route is present in the BGP RIB, this route will be advertised to the neighbor. When a default route is not present in the BGP RIB, one will be artificially generated and advertised to the neighbor for which this feature is configured. This artificially generated default route will not be installed in the BGP RIB. The default-route ( existing in BGP RIB or artificially generated )...
Continue reading →

Configurable log archiving destination

Description Log archiving is used to periodically (every minute) copy logs and core files from tmpfs filesystems /var/log and /var/core to a persistent destination. It was introduced in EOS-4.15.0F and used the external SSD mounted in /mnt/drive as a destination for archiving logs. This feature enhances the log archiving capabilities to be able to use the flash instead of the SSD drive for archiving logs and to configure the destination using the cli. Logs are periodically copied to an archive directory named archive in the persistent destination (/mnt/drive or /mnt/flash). The space size that can be used for archived logs...
Continue reading →

Default Admin Status of Ethernet Interfaces

Description This feature sets the default administrative status of physical Ethernet interfaces to “down/disabled” (e.g. “interface Ethernet1/1; shutdown”). By default, upon creation, interfaces are in the “no shutdown” state (“administratively up/enabled”). In some cases that might be undesirable, such as when a switch is mounted into a data center and cabled prior to it being configured by some management system.  This feature guarantees that a switch will not start to send/receive traffic inadvertently when first powered up or after a new line card inserted (“silent until configured”). There are 2 parts to this feature. Enabling it on a running system, and enabling it “at...
Continue reading →

Support for matching AS path length in BGP

Description The route map feature in BGP allows filtering and manipulating BGP path information. It combines set statements with matching on various path attributes, to make these set statements conditional. This feature provides ability to match on AS path length. Using this match statement, BGP path information could be modified or filtered based on the length of AS path associated with it. This feature is available with the multi-agent routing protocol model and the ribd routing protocol model. It is applicable to VRFs as well. The AS path may consist of different segments: AS_SET, AS_SEQUENCE, AS_CONFED_SEQUENCE, AS_CONFED_SET. For the purpose...
Continue reading →

Set BGP Multi-Exit-Discriminator for IGP Routes

This feature allows for setting BGP path attribute Multi-Exit-Discriminator for IGP routes advertised by BGP. This feature supports OSPF, OSPFv3 and ISIS as IGP. By default, no MED value is set on exported routes. To set MED value, configure a route map using the command ‘set metric’ and apply it. This route map can be applied as an outbound route map or when redistributing routes from IGP into BGP. This command only sets the BGP MED on BGP routes which are redistributed from IGP and has no effect otherwise. Platform compatibility This feature is platform independent. Configuration The following example...
Continue reading →

Forwarding Unicast PTP Packets in Boundary Mode

Description Until now, all PTP packets received on Arista switches with PTP boundary mode enabled will automatically be sent to the control plane to be processed. This is irrespective of whether the PTP packet is multicast or unicast. All PTP packets which have either the destination UDP port 319 or 320, or use the PTP ethertype will automatically be sent to the control plane. This release introduces a new feature which can be enabled by configuring “ptp forward unicast” globally on the switch while “ptp mode boundary” is configured. With this feature enabled, multicast PTP packets will continue to be...
Continue reading →

Notification Message Support for BGP Graceful Restart

Description The BGP graceful restart mechanism defined in RFC4724 limits the usage of BGP graceful restart to BGP protocol messages other than a BGP NOTIFICATION message. The new RFC4724 extension [bgp-gr-notification] permits the graceful restart procedures to be performed when the BGP speaker receives a NOTIFICATION message or HOLDTIME expires. The extension also defines a new BGP NOTIFICATION Cease Error subcode whose effect is to request a full session restart instead of a graceful restart. At a high level, the extension can be summed up as follows. When a BGP session is reset by a NOTIFICATION message, both speakers operate...
Continue reading →

BGP IPv6 Link Local Peers Discovery

Description This feature adds support for a dynamic configuration model to eliminate the need for the network administrator to assign and configure IPv6 addresses for BGP peering. We leverage the following details to automatically establish BGP adjacency: IPv6 link local addresses can be generated by the system via MAC addresses. IPv6 router advertisement can be used to obtain these addresses to form potential BGP peers. As this feature uses IPv6 router advertisement to discover the peer’s IPv6 link local address, it is expected that devices should be IPv6 routing enabled and the interface used for peering should have an IPv6...
Continue reading →

Improved Feature Limitation Check for SSU

Description This feature enhances the feature limitation checking when performing Smart System Upgrade (SSU) and enables software to block SSU downgrades. It also adds a new CLI show command for checking upgrade image compatibility prior to standard upgrades and SSU. More information on SSU feature limitations can be found in the SSU page. Platform compatibility Standard upgrade compatibility checks are supported on all Arista switches. SSU compatibility checks are supported on: DCS-7050X (excluding DCS-7050SX-72, DCS-7050SX-96, DCS-7050TX-72 and DCS-7050TX-96) Configuration There is no specific configuration for this feature. Show Commands The “show version compatibility <URL>” CLI command lets users check the...
Continue reading →

Discontiguous IPv4 Mask Algomatch 7160 Series ACL Support

Description On Algomatch platforms, expansions and encoding transformations on IP prefixes are performed before attempting to fit the transformed set of rules into the ACL hash tables. However, previously Algomatch would only accept IP prefixes (both IPv4 and IPv6 variants) with masks that could be represented in integer format (ex. /16, /24,etc. ) and skip any rules that contained invalid IP prefixes with a trace message stating that the rule was unsupported. This limit meant that IP rules that contained discontiguous masks could not be accepted. Discontiguous masks are IP masks that cannot be represented in integer format, such as...
Continue reading →

“show interfaces interactions” CLI command

Description The show interfaces interactions command aims to provide users a resource that explains various relationships between ethernet interfaces. It describes interactions in which a configuration on an interface causes another set of interfaces to become inactive or have reduced capabilities. Examples include a primary interface consuming subordinate interfaces to service a four-lane speed or platform restrictions that require four interfaces of a port to operate at the same speed. Platform compatibility This command is intended to be platform independent. Configuration N/A Show Commands Syntax The syntax command is show interfaces [intf] interactions [speed [speed]] The user will be able to restrict...
Continue reading →

Multi-VTEP MLAG

Description In conventional VXLAN deployments, each MLAG pair of switches are represented as a common virtual VTEP. VXLAN traffic can be decapsulated on either switch. In some networks, there are hosts that are singly connected to one of the MLAG pair. VXLAN packets destined for the singly connected host could land on the other MLAG peer and subsequently be forwarded over the MLAG peer-link to reach the destination host. This path is undesirable since it would use up some bandwidth on the peer-link. Figure 1 : Suboptimal Packet Path to Singly Connected Host The Multi-VTEP MLAG feature prevents unicast VXLAN...
Continue reading →

Support for new OpenConfig paths

Description 4.21.3F release supports reading and streaming various OpenConfig configuration and state models over gNMI (gRPC Network Management Interface), RESTCONF, and NETCONF transports. A subset of the configuration models may also be modified over these transports, see below. All client transactions that modify device configuration provide the same atomicity guarantees that are provided by sessions in the CLI. Platform compatibility All Configuration The following section outlines configuration options for OpenConfig, NETCONF, and RESTCONF transport methods. Native OpenConfig CLI gNMI Transport This starts a gRPC server listening on address TCP/6030. This server supports both openconfig.proto and gnmi.proto. See gRPC section for...
Continue reading →

EOS SWI image verification

Description SWI Image Verification is a feature that allows one to determine if their SWI image has been tampered with. It verifies that the SWI image was authorized by Arista Networks for release and unaltered post-release. To achieve this, starting from EOS-4.21.3F, a cryptographic signature has been added to Arista EOS images. The cryptographic signature has two major components: a signing certificate, and a signature that was signed by the signing certificate. The signing certificate is trusted by a root certificate stored in the SWI, by default Arista’s root certificate. Together, this information allows one to determine if their SWI image...
Continue reading →

MMU Profile Reservation in Cells

Description MMU Profile CLI allows to reserve buffers in either cells or bytes. This feature eases the configuration as different platforms has different cell size. Platform compatibility 7050’s 7250’s 7300’s 7260’s 7060’s All platforms. Configuration Queue Profiles To configure the MMU parameters for a queue the user needs to create a queue profile. This can be done as follows : switch(config)# platform trident mmu queue profile <name> switch(queue-profile-<name>)# This would create and save a new queue config profile with name set to <name>. If a queue config profile with the given name already exists then entering this mode will allow...
Continue reading →

FPGA repair on CRC error

Description Some of Arista’s products may use one or more FPGAs for certain tasks pertaining to controlling one or more of the following: forwarding ports, LEDs, power supplies, etc. The “FPGA repair on CRC error” feature provides a new option “repair” as an action when an FPGA encounters an error. When the “repair” option is set, the FpgaAgent will attempt to repair the FPGA without any impact (referred to as being “hitless” repair). The sequence of hitlessly repairing the FPGA involves several major steps: Several EOS agents will be shut down The FPGA will be hitlessly reset EOS agents that were shut...
Continue reading →

FPGA repair during ASU

Description All modern high-density electronic circuit chips are susceptible to naturally occurring phenomenon called single event upsets (SEUs) which can randomly flip bits in the chip. Some chips are immune by design, some chips have a built-in detection circuit, while some others may include automatic correction as well. A detection circuit may calculate a CRC on the logic in the chip and assert a CRC error if it detects a bit flip has occurred due to an SEU. When a CRC error occurs, in most cases the device continues to operate as configured. This is because the error may occur...
Continue reading →

Virtual IP Failover Support

Description In certain VXLAN deployments with either NSX/Logical Router or EVPN IRB, active nodes hosting a virtual IP address (VIP) are checked (by standby nodes)  to see if the VIP still exists by broadcasting an ARP broadcast request with VIP as the target.  This is similar to how DHCP will ARP and ping an address before leasing it out.  In EVPN with ARP proxying, the VXLAN software client will use its ARP cache to answer the ARP instead of allowing it to go unanswered and thereby defeating the purpose of the above check and a potential failure goes undetected. Furthermore,...
Continue reading →

Two rate three color marker(TrTCM)

Description 4.21.3F introduces support for the feature on the platforms listed below. The TOI describing the feature support on earlier versions/platforms is available here. Platform capability 7050SX3-48YC12 7050CX3-32S 7300X3-32C-LC 7300X3-48YC4-LC Supported on 7050SX3-48YC12 from 4.21.3F Additionally, action ‘set dscp’ will be supported for yellow packets Configuration Example: 7060X(config)#policy-map type qos PMAP1 7060X(config-pmap-qos-PMAP1)#class CLASS1  7060X(config-pmap-c-qos-pmap1-CLASS1)#police rate 2680000000 bps burst-size 1024 kbytes action set dscp <xyz> rate 5360000000 bps burst-size 1024 kbytes

SNMP MIB for MMU Watermark Counters

Description The MMU watermark counters report the maximum usage of the MMU queues during a given time window. This feature reports the values of those watermark counters for each interface on the system through SNMP.  The values are updated every 30 seconds, and will reflect the higher MMU queues usage during the 30 seconds window immediately preceding the update. Platform compatibility DCS-7050X DCS-7060X DCS-7260X DCS-7300X DCS-7320X Configuration The reporting is enabled with the command Arista(config)#platform trident mmu queue watermark mib It is disabled with the command Arista(config)#no platform trident mmu queue watermark mib The MIB name is ARISTA-XGS-MIB.txt. Limitations This...
Continue reading →

Follow

Get every new post on this blog delivered to your Inbox.

Join other followers: