CONVERGED INFRASTRUCTURE
service or data stream at a time and so is unable to give a realistic test scenario for today’s multi-service carrier Ethernet connections - which often have multiple VLANs within VLANs (also known as QinQ) and MPLS to give the necessary quality of differentiation for different traffic types. This means that RFC2544 is a bit
like testing a motorway with just cars, just lorries or just fire engines - but impossible to test with all of these at the same time. This is not very realistic. The third drawback is that the
standard can actually only test one parameter of one service at a time – so it is an extremely long test process.
Y.1564 The carrier Ethernet sector has been busy developing a new test regime, with the usual snappy sort of name – Y.1564. This allows testing of multi-stream services all at the same time together with nested VLANs, MPLS and many more essential bells and whistles. Not only does Y.1564 mimic real-life
for the network under test, the fact that it can permit testing of all the streams simultaneously means that testing time is cut drastically compared to the old RFC2544 method. Once all configuration errors have
been identified and rectified and the Service Configuration Test has been
passed, Y.1564 moves on to the Service Performance Test. Here the long-term performance
of the Ethernet service is tested with simulated `real traffic’ for all of the services, nested VLANs and MPLS set- ups concurrently. This tests frame transfer delay (FTD), frame delay variation (FDV), information rate (IR), frame loss ratio (FLR) and availability (AVAIL) on each service whilst operating concurrently. Run for between 15 minutes
(minimum recommended) and 24 hours, the Service Performance Test is a highly realistic simulation of live traffic and tests the traffic policing conditions - proving whether the SLA conditions are being met.
SLA policing The fact that Y.1564 testing fully proves whether or not the carrier’s SLA is being met means that a Y.1564 tester like our UniPRO MGig1 is a very useful tool in the armoury of enterprise and data centre network managers - since it is exactly the type of instrument the carriers themselves use to test leased lines. And by the way, carriers will simply ridicule measurements taken with two PCs or two FTP servers over a WAN circuit as being hopelessly inaccurate. The `top ten bandwidth users’ test is
proving to be highly popular as it often avoids a protracted SLA dispute and allows network managers to re-schedule and smooth traffic to get the best from expensive WAN circuits. As any engineer who has ever had
to test Carrier Ethernet links for service turn-up will know, its very rare they are configured correctly first time and so in practice getting a link to pass the Service Configuration Test can take many wasted hours whilst trial-and-error investigations are performed to find out which patch-cord is mis-patched; or which switch or router has been mal- configured. So in our own UniPRO MGig1 for
example we’ve added ‘Isolated Target Check’ and ‘Individual Service Path Set- up - to separate target, service and test items prior to Y.1564 testing. Being able to separate out these
elements during pre-test troubleshooting typically saves many man-hours per link test.
www.idealnetworks.co.uk
Carrier Ethernet – testing methodology 30 NETCOMMS europe Volume IV Issue 1 2013
www.netcommseurope.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44