Compare Products

Hide

Clear All

VS

Time: November 18th, 2024

Preface

Have you experienced similar challenges in the deployment of data centre projects?
During the implementation of Access Control List (ACL)-related services on a data centre switch, one may encounter scenarios where only a limited number of ACL entries are configured, yet the switch indicates that the Ternary Content Addressable Memory (TCAM) resources are insufficient, preventing the configuration from being applied.
There exists a variety of TCAM modes available for data centre switches, and it is often unclear which mode is most appropriate for the current deployment environment.
Furthermore, when planning service equipment, it remains ambiguous what factors should be taken into account to guarantee that TCAM resources are adequately provisioned.

This article aims to address the aforementioned issues. Additionally, it endeavours to enhance your understanding of the underlying principles governing the allocation of TCAM resources to ACLs, thereby facilitating the planning of services within your data centre network and mitigating the risk of inadequate TCAM resources.


Introduction to TCAM

TCAM, or Ternary Content Addressable Memory, serves as a storage solution for Access Control List (ACL) entries within data centre switch products. The resources of TCAM are typically categorized into three types based on the location of the ACL application: VFP ACL TCAM resources, IFP ACL TCAM resources, and EFP ACL TCAM resources. The fundamental unit for storing ACL entries within TCAM is termed a "Slice," with each Slice possessing the capacity to store multiple ACL entries, contingent upon its designated capacity.

TCAM Classification
Application location
VFP ACL TCAM.
Used for QinQ and other scenarios to control outer VLAN changes.
IFP ACL TCAM.
Inbound ACL: perform access control in the inbound direction.
EFP ACL TCAM.
Outbound ACL: perform access control in the outbound direction.

Single Slice (single width) and Double Slice (double width)
A slice of a data centre switch possesses a fixed bit width, although this width may vary slightly among different switch models. For instance, the bit width of each slice in the S6510-48VS8CQ model is 160 bits. When the bit width of a service does not exceed that of a single-width slice, it can be accommodated within one slice, which is referred to as utilizing a single slice. Conversely, if the service's bit width exceeds that of a slice, it necessitates the use of two adjacent slices (beginning with an even number) to create a double slice that provides a greater bit width. In this scenario, the service is classified as utilizing a double slice.

The bit width of a service is determined by calculating the total of the matching fields present within the Access Control Entries (ACEs) in the Access Control List (ACL). This total is then compared to the bit width of the slice. If the calculated bit width of the service does not surpass the bit width designated for a single-width slice, the service will utilize the single slice. Conversely, if the bit width of the service exceeds that of the single-width slice, the service will be assigned to the double slice.

Matching Domains
Illustration
Space occupied (bits)
Vid
Resources required for application on SVI interface or Layer 3 sub-interface
16
Etype
Matching message-type resources
16
Protocol
Resources matching IP protocol numbers
8
Port
Resources matching ports
8
PortGroup
Port reuse resources
12
IPv4 Src
Source IPv4 Resource
32
IPv4-Dst
Destination IPv4 Resource
32
IPv6_Src
Source IPv6 Resources
64 (prefix length is less than or equal to 64) or 128 (prefix length is greater than 64)
IPv6 Dst
Destination IPv6 Resource
Rangcheck
Match resources in the Layer 4 port number range
32
L4_Sport
Layer 4 source port
16
L4_Dport
Layer 4 destination port
16
TCP_Flag
Resources matching TCP Flag
8

The Etype, Port, and PortGroup matching fields are provided based on the specific requirements of the business application, and the Access Control List (ACL) does not have any visual interface for display purposes.

TCAM Features of Box Switches and Frame Switches
The TCAM resource characteristics of various data centre switch models exhibit slight variations and can be broadly categorized into two types:

For box switches, the TCAM resources are classified into three distinct categories: VFP ACL TCAM resources, IFP ACL TCAM resources, and EFP ACL TCAM resources. These resources are allocated independently and are not sharable. Within the IFP ACL TCAM resources of box switches, slices are categorized into two types based on their usage: ordinary slices and reserved slices. Ordinary slices can be combined into double slices by pairing two adjacent slices (starting with even-numbered slices) to accommodate services requiring longer matching domains. In contrast, reserved slices do not support splicing; they can only operate in single-slice mode and are designated for specific services or services planned under certain TCAM configurations. Services utilizing reserved slices are restricted to a single slice (single width). Typically, the initial N slices are designated as reserved slices, where N represents one-third of the total number of slices (for instance, the IFP ACL of the box switch RG-S6510-48VS8CQ comprises 12 slices, four of which are reserved—namely, Slice0 through Slice3). Currently, VFP ACL TCAM resources and EFP ACL TCAM resources do not include any reserved slices; all available resources in these categories are classified as common slices.

● In the case of modular switches, the TCAM resources are divided into IFP ACL TCAM resources and EFP ACL TCAM resources. The key distinction between modular switches and box switches lies in the shared and allocated nature of the IFP ACL TCAM resources and EFP ACL TCAM resources in modular switches. Modular switches possess two small slices (the first two slices) classified as reserved slices. Unlike box switches, these reserved slices can be combined, thereby supporting both Single Slice (single-width) and Double Slice (double-width) services, albeit with specific usage restrictions. The reserved slices are permitted for use solely with designated services.

Example of box switch S6510-48VS8CQ TCAM resource Example of a box switch N18000-X TCAM resource


TCAM Resources

Data centre switches play a crucial role in enabling various services that utilize Access Control Lists (ACLs). It is essential to recognize that not all of these services make use of Ternary Content Addressable Memory (TCAM) resources. For example, the ACL implemented for Virtual Terminal Lines (VTY) to regulate device login does not engage TCAM resources. The following table outlines common services that employ TCAM resources for your reference.

Common services using TCAM resources
The service name displayed on the device
Use ACL on the interface
SECURITY / SECURITY(V6)
Use ACL globally
SECURITY-GBL/SECURITY-GBL(V6)
Traffic Statistics (Counter-only) Service
SECURITY-COUNT/SECURITY-COUNT(V6)
Policy Routing Service
PBR /PBR(V6)
ERSPAN Business
PBM / PBM(V6)
QOS Service
QOS/Q0S(V6)/QOS-CAR/QOS-CAR(V6)
INT Business
INT-INGRESS /INT-INGRESS(V6)
VXLAN Services
VXLAN-CPP/VXLAN-CPP(V6)
CPP Business
CPP/CPP(v6)

Different services utilize TCAM resources based on the Slice dimension. While the same service may access the same Slice resource, it is generally the case that most distinct services cannot share the same Slice resource. Consequently, in typical situations, a given Slice can accommodate only one type of service. For instance, if a Slice is already allocated for SECURITY services (where Access Control Lists are employed under the interface), it cannot concurrently store resources for Policy-Based Routing (PBR) services. Therefore, PBR services must be allocated to a separate Slice, even if the Slice designated for SECURITY services has available capacity. Certain services can be integrated in a specific TCAM mode, allowing these integrated services to occupy the same Slice resources; however, this arrangement is subject to specific priority restrictions.

For the CPP service, the aforementioned services will only utilize TCAM resources after the service has been configured. The CPP service is designed to protect the device's CPU through a standard Quality of Service (QoS) differential service model. It is enabled by default upon device startup and cannot be disabled. As a result, when the device is initiated without any configuration, the CPP service automatically consumes TCAM resources by default.


TCAM Operation and Maintenance Method

TCAM Usage in Data Centre Switches
In a data centre switch environment, the command "show acl res" is employed to display a summary of slice resource utilization. This includes essential information such as the total number of slices, the capacity assigned to each slice, and an overview of current usage along with the remaining resources. For a more comprehensive examination, the command "show acl res detail" can be utilized. This command provides detailed insights into the allocation of slice resources, enabling users to identify which services are utilizing each slice, whether the allocation is single or dual slices, the priority assigned to each service, and other pertinent details.

For instance, to assess the TCAM resource utilization of the RG-S6510-48VS8CQ
switch, one may execute the "show acl res" command. The output reveals that the VFP ACL TCAM resources comprise four slices (Slice 0 to Slice 3), each with a capacity of 256, resulting in a cumulative total capacity of 1024. The IFP ACL TCAM resources consist of twelve slices (Slice 0 to Slice 11), which are divided into four reserved slices (Slice 0 to Slice 3) and eight common slices (Slice 4 to Slice 11), with each slice possessing a capacity of 768, leading to an overall capacity of 9216. Furthermore, the EFP ACL TCAM resources also consist of four slices (Slice 0 to Slice 3), each with a capacity of 512, culminating in a total capacity of 2048.



The detailed utilization of TCAM resources for the fixed switch RG-S6510-48VS8CQ can be examined by executing the command "show acl res detail." The acquired information is as follows: Within the IFP ACL TCAM resources, the CPP service occupies a Double Slice resource, utilizing Slices 10 and 11, while the SECURITY service occupies a Single resource, utilizing Slice 4. Regarding the EFP ACL TCAM resources, the SECURITY service occupies a Single resource, utilizing Slice 0.



To assess the TCAM status of modular switches designated RG-N18000-X,
the "show acl res" command is employed. The information obtained indicates that each board utilizes both IFP ACL and EFP ACL TCAM resources, encompassing a total of 14 slices (Slice 0 through Slice 13). This configuration comprises 2 reserved slices (Slices 0 and 1), each with a total capacity of 128, alongside 12 common slices (Slices 2 through 13), each possessing a total capacity of 2048, resulting in an overall capacity of 24832.



The specific usage of TCAM resources for the modular switch RG-N18000-X can likewise be reviewed by issuing the command "show acl res detail." In this instance, the CPP service occupies one Double Slice resource, utilizing the reserved slice spaces of Slices 0 and 1. Furthermore, the CPP (IPv6) service occupies one Double Slice resource, employing Slices 10 and 11, and the SECURITY (IPv6) service utilizes another Double Slice resource, making use of Slices 12 and 13.



Concerning viewing and adjusting the TCAM mode of data centre switches

It is important to note that, as previously stated, most distinct services are unable to share the same Slice resource. The factory TCAM mode for data centre switches represents the default configuration. In this mode, TCAM services are generally not integrated, and different services are precluded from utilizing the same Slice resource. Typically, a service necessitates the allocation of a minimum of one to two Slice resources. The TCAM mode facilitates a mechanism for selecting the appropriate TCAM operating mode based on various scenarios, allowing certain services to share the same Slice resource or utilize reserved Slice resources. This reduction in Slice utilization for specific functions ultimately increases the availability of Slice resources. For instance, in the ACL-MAX mode employed by certain switch models, both the SECURITY service and the SECURITY-COUNT service can be integrated, permitting the shared use of the same Slice resource.

TCAM Mode
Suitable for use scenarios
Default mode
Suitable for common scenarios with few TCAM service types
ACL-MAX Mode
Suitable for IPv4/IPv6 dual-stack scenarios
ACL-SLICES-ALL Mode
A special mode of ACL-MAX mode, which increases the security ACL capacity based on ACL-MAX mode. Only some models of devices support it.
VXLAN Mode
Suitable for VXLAN scenarios
VXLAN-HASH Mode
Suitable for VXLAN scenarios that support VXLAN hash load balancing Suitable for general scenarios that support VXLAN hash load balancing
ACL-DEFAULT-VXLAN-HASH Mode
Suitable for use scenarios

To examine the TCAM working mode of the data centre switch, one may utilize the command "show team-mode status." This command provides insights into the current TCAM mode of the device as well as the TCAM mode that will be in effect following a reboot of the device. Should there be a need to modify the TCAM mode, the command "team-mode acl-max/acl-slices-all/vxlan/vxlan-hash/acl-default-vxlan-hash" can be employed for adjustments. It is important to note that after altering the TCAM mode, it is necessary to save the configuration and perform a complete restart of the device for the changes to take effect.



TCAM Application and Practice

TCAM Resource Assessment Recommendations

Data Centre Switch TCAM Resource Evaluation Factors
Slice Resource Evaluation: When assessing slice resources, it is imperative to evaluate whether the slice capacity accommodates the access control list (ACL) entries. However, it is of greater importance to determine if the number of available slices can meet the service requirements. As noted previously, in most instances, a single slice is capable of storing only one type of service. In practical applications, it is possible to encounter situations where there is surplus slice capacity, yet no additional complete slices are available for new services. This scenario is likely to result in insufficient TCAM resources and warrants careful consideration.
KEY Resource Evaluation: The KEY refers to a processing engine within the device that allocates TCAM resources to services necessitating such resources. Box switches do not require a separate evaluation of KEY resources; an evaluation of Slice resources alone suffices. If Slice resources are adequate, they will also inherently satisfy KEY resource requirements. However, it is essential that switches equipped with DUNE series chips conduct a separate assessment of KEY resources, particularly for those operating on version 12. X.

Method for Evaluating TCAM Resources in Data Centre Switches:
1. Assessing Slice Capacity
Against ACL Entries: Given the diverse configurations of access control entries (ACEs) within ACLs, the slice capacities employed by various ACEs differ and cannot be predetermined. If a substantial number of ACEs are implemented within the ACL, it is advisable to configure the deployed services on the specific data centre switch model to verify whether the slice capacity is adequate for the ACL entries.
2. Determining Sufficiency of Slices for Business
Needs: First, compare the bit width associated with the services described in prior sections against the bit width of the slices to ascertain whether each service necessitates a single or dual slice. Next, aggregate the slices required by all services to derive the "total number of slices required for the service." Finally, evaluate whether the total number of slices aligns with the service requirements, referencing the accompanying table.
When calculating the "total number of slices required for the service," it is crucial to account for the slice resources utilized by the device's default Common Power Policy (CPP) service within the IPv4 context. Similarly, in a dual IPv4 and IPv6 scenario, one must also consider slice resources employed by both the default CPP service and the CPP (v6) service.

EFP ACL Assessment for Box Switches
Evaluation conditions
Evaluation results
Required for outbound services is less than the number of EFP ACL slices.
Slice has sufficient resources.
The number of EFP ACL slices is less than the number of slices required for outbound services.
Adjust TCAM mode or integrate used services
IFP ACL Evaluation for Box Switches
Evaluation conditions
Evaluation results
The number of slices required for inbound services is less than the number of IFP ACL common slices.
Slice has sufficient resources.
The number of IFP ACL common slices < the number of slices required for inbound services < the total number of IFP ACL slices
Configure the deployed services to the actual data centre switch model for verification.
The total number of IFP ACL slices is less than the number of slices required for inbound services.
Adjust TCAM mode or integrate used services.

Frame switch evaluation
Evaluation conditions
Evaluation results
The total number of slices required for inbound and outbound services is less than the number of IFP/EFPACL common slices.
Slice has sufficient resources.
The number of IFP/EFP ACL common slices < the total number of slices required for inbound and outbound services < the total number of IFP/EFP ACL slices
Configure the deployed services to the actual data centre switch model for verification.
The total number of IFP/EFP ACL slices is less than the total number of slices required for inbound and outbound services.
Adjust TCAM mode or integrate the services used.

3. Please evaluate whether the KEY resources meet the requirements:
In the context of box switches, it is not required to assess KEY resources independently; rather, only slice resources warrant evaluation. This section pertains exclusively to box switches. When assessing KEY resources, please refer to the quantity of KEY resources as indicated in the accompanying table. Each KEY resource should be treated as equivalent to one Slice, and the evaluation should proceed in accordance with the established methodology for "Evaluating whether the number of Slices can meet the business requirements" outlined in the preceding section. Generally, in scenarios involving both IPv4 and IPv6, should there be a deficiency in KEY resources, it is advisable to modify the TCAM mode to ACL-MAX mode. In this mode, allocations for IPv4 and IPv6 KEY resources occur separately, thereby increasing the total available KEY resources.

The default mode of modular switches
KEY Resource Type
KEY Resource Type
IFP ACL KEY resource
7
EFP ACL KEY resource
2
ACL-MAX Mode for Module Switches
KEY Resource Type
KEY Resource Type
IFP ACL IPv4 KEY resource
7
IFP ACL IPV6 KEY resource
7
EFP ACL KEY resource
2

4. Case Study on TCAM Resource Insufficiency
This document provides an analysis of real-world instances of insufficient TCAM resources, aimed at enhancing the understanding of the underlying logic involved in TCAM resource allocation for Access Control Lists (ACLs). In situations where TCAM resources are insufficient, it is advisable for practitioners to first consider adjusting the TCAM mode in alignment with the specific business scenario. Upon adjusting the TCAM mode, the device will autonomously optimize the TCAM resources for the most frequently utilized services in that scenario. Following this modification, it is essential to assess whether the TCAM resource requirements of any new services can be fulfilled.

Should the issue of insufficient TCAM resources persist after the mode adjustment, the next recommended step is to consolidate the currently employed TCAM services, followed by implementing necessary adjustments or deletions.


Example: Insufficient Slice Resources in RG-S6510-48VS8CQ Switch
The subsequent TCAM service is deployed on an RG-S6510-48VS8CQ within an IPv4 context:

Business Name
Business Configuration
Equipment business name
TCAM Resource Type
CPP (IPv4 )
CPP
IFP
Use ACL on the interface (IP4)
ip access-list extended IPV4 ACL IN
SECURITY
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
p access-list extended IPy4 ACL OUT
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
interfaceTFGigabitEthernet 0/1
AFP
ip access-group IPv4 ACL IN in
ip access-group IPV4 ACL OUT out
Traffic Statistics (Counter-only) (IPV4)
ip access-list extended COUNT IN
SECURITY-COUNT
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
20 permit ip any any
Ip access list extended COUNT OUT.
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eg 2222
20 permit ip any any
AFP
interface TFGigabitEthernet 0/2
ip access-group COUNT IN in counter-only
ip access-group COUNT OUT out counter-only
Global ACL (IPv4)
ip access-list extended GBL IN
SECURITY-GBL
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
20 permit ip any any
ip access-group GBL IN in
Policy Routing (IPV4)
ip access-list extended PBR IPv4
PBR
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
route-map PBR IPv4 permit 10
Match IP address PBR IPv4.
Set ip next-hop 3.3.3.3
interface TFGigabitEthernet 0/3
ip policy route-map PBR IPv4
QOS (IPV4 )
mac access-list extended ARP NO FORWARDING
QOS-CAR
AFP
10 permit any arp
class-maP ARP_NO_FORWARDING
Match access-grouP ARP NO FORWARDING.
policy-maP ARP NO FORWARDING
class ARP NO FORWARDING
police 00exceed-action drop
service-policy output ARP NO FORWARDING
QOS (IPV4 )
expert access-list extended RoCEv2-Cnp
QOS
IFP
10 permit udp any any any any eq 4791 dscp 35
Expert access-list extended RoCEv2-Data.
10 permit udp any any any any eq 4791 dscp af41
class-map RoCEv2-Data
match access-group RoCEv2-Data
class-map RoCEv2-Cnp
match access-group RoCEv2-Cnp
policy-map RoCEv2
class RoCEv2-Data
Set cos 1 priority
class RoCEv2-Cnp
Set cos 2 priority
service-policy input RoCEv2
ERSPAN
expert access-list extended RDMA
PBM
IFP
10 permit udp any any any any eq 4791 udf 1 l5 _head 0 0x8100 0xFF00
20 permit udp any any any any eq 4791 udf 1 15 head 0 0x1100 0xFF00udf 2 15 head 12 0x6000 0xFF00
Monitor session 1 span-source
source interface HundredGigabitEthernet 0/56 rx acI RDMA
original IP address 1.1.1.1
destination IP address 2.2.2.2

Upon the completion of the service deployment, it is essential to verify the TCAM (Ternary Content Addressable Memory) resource utilization on the device. The accompanying table illustrates that the IFP (Ingress Filter Policy) ACL (Access Control List) TCAM resource consists of 12 slices, which include 4 reserved slices and 8 common slices. Currently, 1 reserved slice and 6 common slices are being utilized. In terms of the EFP (Egress Filter Policy) ACL TCAM resource, there are 4 slices available, of which 3 slices are currently in use.



The subsequent figure delineates the specific slice usage by services. For IFP ACL TCAM resources, the QOS (Quality of Service), PBR (Policy-Based Routing), PBM (Policy-Based Management), and SECURITY-COUNT services utilize individual slices, with each consuming one slice. The SECURITY services and SECURITY-GBL services collectively share one slice, utilizing it in total. Additionally, CPP (Class packet processing) services require two slices, as they function as double slices. Regarding the EFP ACL TCAM resources, the SECURITY services, SECURITY-COUNT services, and QOS-CAR services each employ single slices, utilizing one slice each.



In light of the services outlined above, the device is required to activate the IPv6 feature and implement the IPv6 interface ACL function to accommodate specific business requirements. The detailed deployment functions are as follows:

Business Name
Business Configuration
Business Configuration
TCAM Resource Type
Enable IPv6 configuration
interface VLAN 100
CPP(v6)
IFP
ipv6 enable
ipv6 address FC01::1/64
ACL Interface (IPv6)
ipv6 access-list IPv6 ACL IN
SECURITY(v6)
IFP
10 permit icmp host FC01::1 host FC01::2
ipv6 access-list IPV6_ACL OUT
10 permit icmp host FC01::l host FC01::2
interfaceTFGigabitEthernet 0/1
AFP
ipv6 traffic-filter IPv6_ACL_IN in
ipv6 traffic-filter IPv6_ACL_OUT out

Following the activation of the IPv6 functionality, the CPP (v6) service is configured to utilize the Double Slice IFP ACL TCAM resources by default, with an expected reliance on common Slice 8 and Slice 9. Subsequently, upon configuring the IPv6 interface ACL service and invoking the ACL in both the inbound (IN) and outbound (OUT) directions at the interface level, the device is likely to encounter inadequate TCAM resources. At this juncture, it is important to note that three reserved Slice resources remain within the IFP ACL TCAM. As the SECURITY (v6) service is categorized as a Double Slice, it is unable to utilize the reserved Slice resources. Additionally, the EFP ACL TCAM currently has one Slice resource available; however, the SECURITY (v6) service requires two Slice resources due to its Double Slice designation. Consequently, both the IFP ACL TCAM and EFP ACL TCAM will experience insufficient Slice resources.

The device initially operated under an IPv4 single-stack configuration. Upon the activation of the IPv6 functionality and the subsequent deployment of a dual-stack, the issue of insufficient TCAM resources emerged. It is advisable to review the recommended TCAM modes for implementation in the aforementioned scenarios. For an IPv4/IPv6 dual-stack configuration, employing the ACL-MAX mode is strongly recommended. Following the alteration of the device's TCAM mode and a subsequent restart, the device commenced operation in ACL-MAX mode. Upon re-enabling the IPv6 functionality and deploying the IPv6 interface ACL function, there emerged no further TCAM resource constraints. At this point, the TCAM resources of the device should be assessed, as depicted in the accompanying figure. The CPP (v6) service and SECURITY (v6) service are shown to be utilizing TCAM resources effectively.



The accompanying figure illustrates the specifics regarding the slices allocated to the various services. For IFP ACL TCAM resources, the QOS services, CPP (v6) services, PBM services, and PBR services each employ one slice as a single slice. In contrast, the SECURITY services, SECURITY-GBL services, and SECURITY-COUNT services collectively utilize one slice. The SECURITY (v6) services and CPP services, being Double Slices, each require two slices. With respect to EFP ACL TCAM resources, QOS-CAR services are single slices that utilize one slice; SECURITY services and SECURITY-COUNT services again share a single slice for total utilization of one slice; whereas the SECURITY (v6) services necessitate two slices as Double Slices.



In the case of insufficient KEY resources within a modular switch (RG-N18000-X), the following TCAM services are deployed in an IPv4/IPv6 dual-stack scenario:

Business Name
Business Configuration
Equipment business name
TCAM Resource Type
CPP (IPv4)
CPP
IFP
Enable IPv6
interface VLAN 100
CPP(v6)
IFP
ipv6 enable
ipv6 address FC01::1/64
Global ACL
ip access-list extended IPv4_ACL IN
SECURITY-GBL
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
20 permit ip any any
ip access-group IPv4_ACL_IN in
Global ACL (IPv6)
ipv6 access-list IPv6 ACL IN
SECURITY-GBL(v6)
IFP
10 permit icmp host FC01::l host FC01::2
ipv6 traffic-filter IPv6_ACL_IN in

Upon the completion of the service deployment, it is essential to assess the utilization of TCAM resources on the device. The findings indicate that the IFP ACL TCAM and EFP ACL TCAM resources share a total of 14 slices, which comprise 2 reserved slices and 12 common slices. At present, 2 reserved slices and 5 common slices are utilized. In addition to the evaluation of slice resources, it is imperative for modular switches to also examine the usage of KEY resources. Currently, there are 7 IFP ACL TCAM KEY resources, with 7 actively in use, and 2 EFP ACL TCAM KEY resources, none of which are currently utilized.



The default mode of modular switches
KEY Resource Type
KEY Resource Type
IFP ACL KEY resource
7
EFP ACL KEY resource
2

Based on the above services, the device adds the IPv4 ERSPAN function. The specific deployment functions are as follows:

Business Name
Business Configuration
Equipment business name
Equipment business name
ERSPAN (IPv4)
ip access-list extended ERSPAN IPv4
PBM
IFP
10 permit tcp host 1.1.1.1 eq 1111 host 2.2.2.2 eq 2222
Monitor session 4 span-source
source interface TenGigabitEthernet 2/1 rx aCI ERSPAN IPV4
original IP address 1.1.1.1
destination address 2.2.2.2

Upon the completion of the service deployment, it is essential to assess the utilization of TCAM resources on the device. The current findings indicate that the IFP ACL TCAM and EFP ACL TCAM resources collectively share 14 slices, comprised of 2 reserved slices and 12 common slices. At present, 2 reserved slices and 5 common slices are actively in use. In addition to analyzing slice resources, it is also imperative for modular switches to evaluate the allocation of KEY resources. Currently, there are 7 IFP ACL TCAM KEY resources, of which 7 are in active use; further, there are 2 EFP ACL TCAM KEY resources, with none currently in use.



ACL-MAX Mode for Module Switches
KEY Resource Type
KEY Resource Type
IFP ACL IPV4 KEY resource
7
IFP ACL IPv6 KEY resource
7
EFP ACL KEY resource
2


Summary

This article is structured into three chapters and seven sections. It presents an evaluation method for the TCAM resources utilized in ACL services for Ruijie data centre switches, examined through three key dimensions: an introduction to TCAM, methods for TCAM operation and maintenance, and its application and practical implementation. The theoretical framework discussed herein applies to a wide range of data centre switches; however, it is important to note that there may be variations across different equipment models. Due to spatial limitations, a comprehensive discussion of all aspects cannot be accommodated. Readers need to acknowledge this limitation. Consequently, when assessing the utilization of TCAM resources, it is strongly recommended that the services being deployed are configured for the specific data centre switch model in use for accurate verification, with the ultimate verification results serving as the definitive reference.



Related Blogs:
Exploration of Data Centre Automated Operation and Maintenance Technology: Zero Configuration of Switches
Technology Feast | How to De-Stack Data Centre Network Architecture
Technology Feast | A Brief Discussion on 100G Optical Modules in Data Centres

Research on the Application of Equal Cost Multi-Path (ECMP) Technology in Data Centre Networks
Technology Feast | How to build a lossless network for RDMA
Technology Feast | Distributed VXLAN Implementation Solution Based on EVPN
Exploration of Data Centre Automated Operation and Maintenance Technology: NETCONF
Technical Feast | A Brief Analysis of MMU Waterline Settings in RDMA Network
Technology Feast | Internet Data Centre Network 25G Network Architecture Design
Technology Feast | The "Giant Sword" of Data Centre Network Operation and Maintenance
Technology Feast: Routing Protocol Selection for Large Data Centre Networks
Technology Feast | BGP Routing Protocol Planning for Large Data Centres
Technology Feast | Talk about the next generation 25G/100G data centre network

Ruijie Networks websites use cookies to deliver and improve the website experience.

See our cookie policy for further details on how we use cookies and how to change your cookie settings.

Cookie Manager

When you visit any website, the website will store or retrieve the information on your browser. This process is mostly in the form of cookies. Such information may involve your personal information, preferences or equipment, and is mainly used to enable the website to provide services in accordance with your expectations. Such information usually does not directly identify your personal information, but it can provide you with a more personalized network experience. We fully respect your privacy, so you can choose not to allow certain types of cookies. You only need to click on the names of different cookie categories to learn more and change the default settings. However, blocking certain types of cookies may affect your website experience and the services we can provide you.

  • Performance cookies

    Through this type of cookie, we can count website visits and traffic sources in order to evaluate and improve the performance of our website. This type of cookie can also help us understand the popularity of the page and the activity of visitors on the site. All information collected by such cookies will be aggregated to ensure the anonymity of the information. If you do not allow such cookies, we will have no way of knowing when you visited our website, and we will not be able to monitor website performance.

  • Essential cookies

    This type of cookie is necessary for the normal operation of the website and cannot be turned off in our system. Usually, they are only set for the actions you do, which are equivalent to service requests, such as setting your privacy preferences, logging in, or filling out forms. You can set your browser to block or remind you of such cookies, but certain functions of the website will not be available. Such cookies do not store any personally identifiable information.

Accept All

View Cookie Policy Details

Contact Us

Contact Us

How can we help you?

Contact Us

Get an Order help

Contact Us

Get a tech support

2024 Ruijie Networks Brand Awareness Survey

Your opinions and feelings are crucial for our improvement.

Fill in the survey