[Home][RAD]

RAD section home tab  Search for RAD equipment  RAD and other WAN / LAN specials  Request a quote on RAD or other WAN / LAN equip.

RAD Data Communications - Authorized Distributor

Service Turn-Up of Carrier Ethernet

Service Turn-Up
Controlled introduction of a new service typically includes such steps as service setup, in-field testing to validate link and service connectivity, and acceptance tests to verify that the service is running smoothly according to the SLA and the QoS (quality of service) and CoS (classes of service) it defines. Testing at this point also serves for generating a baseline for performance parameters, to which future test results will be compared. Specifically, KPI (key performance indicators) metrics are established for end-to-end throughput, packet delivery ratio, latency, and jitter. These results are recorded as the service “birth certificate” and are archived for customer reporting, SLA comparison and future use as needed, for example, to evaluate performance after upgrade or repairs. To accurately establish service behavior over time, burn-in tests are performed for 24-72 hours before the link is cleared for critical applications.

Let us consider, for example, a 3-CoS EVPL (Ethernet Virtual Private Line) service between several locations: Enterprise headquarters, remote branches and a data center. Some of these locations reside outside of the provider’s service footprint and therefore the service path includes segments that traverse another carrier’s network. The provider in this case must ensure that the EVC is operational and running fault-free, that each traffic type adheres to its respective SLA QoS guarantees, and that visibility is established throughout the entire service path, including across a third-party network, so that network operators can accurately pinpoint the origin of problems and ensure a speedy resolution. The following table summarizes the various functions that are typically associated with service turn-up and the different tools network operators use to perform them:
Service turn-up tools and functions for Carrier Ethernet

Further details on selected procedures are provided in the sections below:

Connectivity Verification
Before initiating any tests involving high traffic volumes, it is imperative to verify that the service connections are up and configured properly. Service connectivity is typically verified using 802.1ag/Y.1731 continuity check (CC) messages to test the end-to-end path, whether this path traverses a single or multiple networks.

While other tools may also be used to check that connections are established, the value of CC heartbeat messages is in helping service operators validate that peer elements are defined correctly. This allows quick identification of mismatch errors in MEP configuration – errors which may result in duplicate IDs and service cross-connects between different customers or maintenance domains.

During turn-up connectivity testing, the demarcation device installed at headquarters can be configured to send periodic CC messages to the other units installed at the remote locations and to check for incoming CC messages sent at that period. For each maintenance domain level, i.e., service provider, third-party carrier and enterprise customer, the MEPs are identified by their respective MEP IP, MD (maintenance domain) and MA (maintenance association). Upon receipt of a CCM, the receiver checks the MEP and service IDs to ensure conformance. LOC is declared if no CC messages are received from remote MEPs within 3.5 times the transmission period. Further details on the next steps taken in such cases are provided in Chapter 5: Fault Management and Recovery.

Diagnostic Loopbacks
Ethernet circuit validation at the turn-up stage can be verified with on-demand diagnostic loopbacks. In this case, the demarcation device is configured to loop back tested traffic to the test originator – whether a management station, test head or a demarcation device at another location – at wire-speed. The demarcation device will also swap the MAC/IP source and destination addresses to ensure that the traffic correctly traverses the Layer 2 or Layer 3 network back to the source. This allows network operators to verify service path configuration and to test connection quality for actual traffic, for example by running diagnostic loopbacks for all VLAN IDs mapped to a particular EVC. Other variations of such tests, including multi-flow and “always on” loopbacks, are described in Chapter 5: Fault Management and Recovery.

Stress Testing
RFC-2544 throughput measurements are used to establish the effective bandwidth rate the enterprise actually receives for each tested EVC by determining the maximum traffic load for which no errors occur. The test can be conducted bi-directionally or unidirectionally using a test head device or a demarcation device equipped with built-in RFC-2544 tester capabilities, thus eliminating the need for external testing equipment. The procedure outlines a set of frame sizes for which throughput measurements are conducted. This helps to pinpoint the processing delays caused by short frames and identify network equipment that is having trouble handling larger packets. Based on packet receive-rate results, the service provider can also determine the maximum packet length that can be offered to the enterprise with SLA guarantees, after calculating frame extension allowance for headers, SP-VLAN tags and other overhead. 

As RFC-2544 procedures were originally designed for individual device testing in a lab environment and may be time consuming, performing several tests concurrently is advisable.

SLA Verification
RFC-2544 procedures can also be used to determine network performance and to verify that it is in-line with the KPIs guaranteed in the enterprise SLA. Packet loss is measured by comparing the
number of transmitted and received frames at various network loads and frame sizes, then calculating the percentage of dropped frames. Average round-trip or one-way latency is determined
by analyzing time stamps on tested frames. Packet jitter, or delay variations over time, is also part of the test procedure although, technically, it is not included in the official RFC-2544 standard. A new standard, ITU-T Y.1564 (formerly Y.156sam), has been recently approved to provide a better fit for SLA validation during Carrier Ethernet service activation. It allows efficient measurement of CIR and EIR bandwidth, latency, jitter, and frame loss in a single procedure, using uni- or bi- directional loopbacks between a Y.1564-enabled test unit operating opposite another test unit or opposite a demarcation device with wire-speed support.

Return to Carrier Ethernet with RAD products

 

Immediate price quote - Call 727-398-5252 or e-mail cs@bestdatasource.com

Don't forget to ask about  RADCare support for your RAD products.

Specials   E-mail us   Bookmark this site   More RAD    Download RAD English catalog    Important Notices

Use Visa to purchase Data Communications Equipment    Use MasterCard to purchase Data Communications Equipment
E-mail: Webmaster about the RAD portion this site.
Last modified: February 13, 2021
© 1998-2017 Cutter Networks Inc - All rights reserved