Greeting all,
Is anyone out there using PON in a campus or facility environment? I am talking to a few vendors who are pushing PON as a replacement for edge switching on the campus and in some cases, ToR switch in the DC. Opinions on this technology would be greatly appreciated.
Thanks,
Kenneth
I evaluated several campus PON project on both vendor/provider and customer
sides over the last 5 years.
PON was designed as a last-mile technology for _operators_ to cover huge
areas and save on OPEX - that's the usual vendor's mantra.
GigE usually takes two fibers to deliver; GigE aggregation requires an
Ethernet switch you need to power and maintain - "many points of failure"
argument.
PON only takes one fiber, splitting a PON signal is done by a cheap passive
optical splitter which, once installed, lasts forever.
Correct cost comparison between PON and Ethernet is pretty difficult.
First, the cost of the actual network _equipment_ (switches, routers etc.)
for an Access network (tens of thousands of subscribers) is nothing
compared to the civil works - trenching, cabling, permits etc. Second, CPE
cost is the most sensitive on that scale. $5 cheaper CPE saves more money
than a $50k cheaper OLT.
Without scale benefits and (assumed) OpEx reduction, PON projects are
usually significantly more expensive.
Now, to the projects.
I have never heard of seen PON on a DC level.
As for the campus, none of the projects took off, and here is why:
- PON equipment is proprietary. An OLT (PON hub) from vendor X works only
with ONTs from the same vendor. Every other vendor claims there is "some
interop" but none was able to do demonstrate.
- PON infrastructure - a passive fiber tree - is a) proprietary b) not
flexible. You can make any physical topology of your usual p2p fiber
segments but a tree is always a tree; plus PON usually requires green APC
(angle polished) connectors.
- PON US/DS bandwidth allocation is asymmetric.
Training, maintenance and support are important, too. Ethernet is
ubiquitous, PON engineers are a bit harder to find. Proprietary nature of
the product implies that a Huawei PON would look and feel somewhat
different than, say, Calix PON hence the knowledge is more vendor-specific.
Also, nobody likes vendor lock. If an Access Ethernet switch fails, you can
- potentially - replace it with another fairly easily. PON ONT of different
vendors may all come from the same OEM like Cambridge but the firmware is
different.
What did I miss?... Ah, the usual cabling
There is a more sophisticated WDM-PON. Ericsson and Teradata had it in
development, Infinera offers it as iAccess product - but I haven't tried
that yet.
Overall, it really depends on a project (getting all the MTBF data and
calculating the failure cost was quite a task) but I personally can't see
any PON benefits over Ethernet for either campus or DC even in the midterm.
From my personal experience, flexibility with cables plus cheap CWDM
filters and colored optics can do magic on the campus level.
Stas
Greeting all,
Is anyone out there using PON in a campus or facility environment? I am talking to a few vendors who are pushing PON as a replacement for edge switching on the campus and in some cases, ToR switch in the DC. Opinions on this technology would be greatly appreciated.
The datacenter tor / host evironments I'm exposed to generally consider
10Gb/s or Nx10 to be the performance floor not the ceiling. To my
thinking 100gig lr4 results in a substantial reduction in
fiber/optic/port count which will in the long more than offset the
increased cost something that didn't really happen with 40gig.
Conversely, many places still can't even handle a full table. 
I think it's really depends on your use case. If you know your TOR switches are doing 1-2 gigs at all times PON would be quite expensive to keep up with this type of usage. If your talking distribution/access out to cameras ect it would work. We run gpon across our entire access network 99% of the time it's fine once in awhile you'll get that one user that is killing the pon port and will move them to an active ethernet port. Now when NG-PON2 starts to hit the market, now you're talking about something that could be used in the data center with it's capabilities to do 4X 10G PON bonding. Note this is not going to be cheap but doable.
Carlos Alcantar
Race Communications / Race Team Member
1325 Howard Ave. #604, Burlingame, CA. 94010
Phone: +1 415 376 3314 / carlos@race.com<mailto:carlos@race.com> / http://www.race.com/>
Hi,
using in rural area, it works. Much cheaper than ETTH.
In a message written on Sat, Jan 21, 2017 at 03:22:20PM -0600, Stas Bilder wrote:
Now, to the projects.
I have never heard of seen PON on a DC level.
A friend of mine told me of a fascinating in-data center PON solution.
He had a customer that needed high speed multicast fan-out. They
chose to use a 10G/1G PON solution, a single source node sending
10G PON with a 1G backchannel, 100% of the traffic was multicast.
It could be passively split in DC up to 128:1 with zero "packet
loss", good luck finding an Ethernet switch which can take in 10G
of multicast in and turn it into 1280G of multicast out without
dropping a frame. It was all done entirely inside of a single
data center.
Since then I've mentioned the trick to several other folks I know
who need high speed multicast/broadcast replication. In the DC
distance is rarely an issue, so the solution degenerates to special
SFP's and a splitter, which is pretty dang simple.
However, this is clearly a corner case, and I agree with your
assessment overall.
That's a genious idead