x
<<   <   Page 2 / 4   >   >>
routethus 12/5/2012 | 1:01:50 AM
re: Cisco's CRS-1 Passes Our Test and the CRS-1 was available for testing before the last quarter............

dpb 12/5/2012 | 1:01:50 AM
re: Cisco's CRS-1 Passes Our Test Here is another question no one has even asked: Why, in the quarter after the CRS-1 was announced, did VZ suddenly show up out of nowhere as a 10% JNPR customer. It seems to me that this is not a coincidence, that VZ was waiting on this beast, saw it and went to JNPR. Am I missing something?

Phone companies can't make decisions that fast, you're reading too much into it.

-David Bannister
dpb 12/5/2012 | 1:01:49 AM
re: Cisco's CRS-1 Passes Our Test and the CRS-1 was available for testing before the last quarter............

Which means that the choice to purchase juniper was made at least 9-12 months prior to the booking to show in the quarterly results. In telco land it takes a long time to go from concept, pick hardware, test hardware, do design + specs, backend systems, install hardware and activate hardware. It's a very slow process. Ask anyone that has worked for a telco or has sold to a telco, the typical sales cycle is 12-24 months. Think glacier.

-David Bannister

volkot 12/5/2012 | 1:01:44 AM
re: Cisco's CRS-1 Passes Our Test
That CSR-1 cannot do line rate for small packets
All tests are designed around iMix
Quite an interesting detail
russ4br 12/5/2012 | 1:01:41 AM
re: Cisco's CRS-1 Passes Our Test All tests are designed around iMix

It seems that one of the tests (Software Maintenance) ran with 40-byte packets:

http://downloads.lightreading....

pg.16
"An exception to this rule is the loss-less software upgrade. It uses only 40-byte packets (IPv4) or 60-byte packets (IPv6) because the interrupt time can be measured more exactly with small packets"


As for the results:

http://www.lightreading.com/do...

During the upgrade, the router lost just 24 IP packets on all interfaces, forwarding at a rate of 560 million pps G㢠equivalent to 120 nanoseconds, or 0.00000012 seconds of service interrupt. In the downgrade test, the system lost only 9 packets equivalent to 45 ns


-russ
routethus 12/5/2012 | 1:01:33 AM
re: Cisco's CRS-1 Passes Our Test >> and the CRS-1 was available for testing before the last quarter............

Which means that the choice to purchase juniper was made at least 9-12 months prior to the booking to show in the quarterly results. In telco land it takes a long time to go from concept, pick hardware, test hardware, do design + specs, backend systems, install hardware and activate hardware. It's a very slow process. Ask anyone that has worked for a telco or has sold to a telco, the typical sales cycle is 12-24 months. Think glacier.

-David Bannister <<

This perspective is historically correct. However, carriers are often complaining these days about their management pushing them to make decisions in 3-6 months cycles. BTW, the CRS-1 was available for testing by select carriers, over 9 months ago. What cycle VZ is actually using, I don't know. But let's face it, CRS-1 is an evolving solution. The safe bet today is to go with either the GSR or the T, if you are conservative by nature.

-r
tekweeny 12/5/2012 | 1:01:26 AM
re: Cisco's CRS-1 Passes Our Test All tests are designed around iMix
It seems that one of the tests (Software Maintenance) ran with 40-byte packets:

During the upgrade, the router lost just 24 IP packets on all interfaces, forwarding at a rate of 560 million pps G㢠equivalent to 120 nanoseconds, or 0.00000012 seconds of service interrupt. In the downgrade test, the system lost only 9 packets equivalent to 45 ns


A CRS-1 has a theoretical potential to forward ~1.45bpps, at 40 bytes. They only forwarded 560mpps in the HA tests, or, a little over 1/3rd the "Cisco marketed" capacity.

560mpps forwarding, yawn, that is SO 2002.

Why the pathetically light load during the HA testing? There are competitors that have done this for awhile, using 40B packets, loaded up at line rate, with a similar HA testing setup and zero loss, long before the CRS-1.

Lots of groveling in this report, which just does not jive with the LR I had grown to love over the years. The old LR would have brought the fact that this was a limp middle legged test to light with lots of interesting commentary about how other competitors were (or, claiming) to already be doing this.

Bring back LRs cynacism, and, BobbyMax!
gotman 12/5/2012 | 1:01:25 AM
re: Cisco's CRS-1 Passes Our Test When will you and other take this router and report for face value? From this report it seems like the T640 is poised to loose a lot of bids on a plain level field.(not withstanding discounts)!

The T640 now has a big brother in the core!!
nobody55 12/5/2012 | 1:01:25 AM
re: Cisco's CRS-1 Passes Our Test or it could have nothing to do with core routing at all and boil down to the fact that vz is a large channel for netscreen products.
ragho 12/5/2012 | 1:01:21 AM
re: Cisco's CRS-1 Passes Our Test
or it could have nothing to do with core routing at all and boil down to the fact that vz is a large channel for netscreen products.


You're on the right track, but not quite right. VZ is a big Juniper customer because of the E-series routers (not the M/T/J or NetScreen). VZ is committed to specific FTTx applications and the E-series rules in the broaband aggregation space.

'nuff said. I don't have exact numbers on how the sales numbers were split, but the E-series represents the major chunk.
<<   <   Page 2 / 4   >   >>
HOME
Sign In
SEARCH
CLOSE
MORE
CLOSE