It may depend on what counts as "working". When I had to do a migration, I could get the 3rd party ONT to bridge ethernet traffic, and the OLT could put it in the native VLAN. No configuration, firmware updates, or any other management would work. But this was good enough that we could swap OLT's in a maintenance window and swap ONT's over a period of time.
What is this multi-vendor ONT you found? I'm curious now. ________________________________ From: AF <[email protected]> on behalf of dbernardi <[email protected]> Sent: Tuesday, October 6, 2026 9:59 AM To: [email protected] <[email protected]> Subject: Re: [AFMUG] PON and UPC In my experience ONT/OLT interop is mostly proprietary. A Calix ONT will not work with a Adtran OLT. For GPON and XGS-PON we moved to a ONT vendor that supports multiple OLT vendors by detecting the OLT and changing its operation automatically. For example, we replaced our Calix and Nokia ONTs over time with these more flexible ONTs, then swapped out the OLT later with minimal disruption. The ONT also supports downloading a configuration file via a logical management interface (VEIP/IP-HOST) which can mean far less per subscriber specific OLT configuration. That said, I think very few operators do that, so yes, if you have Calix ONTs in the home, you will need Calix OLT in the headend/cabinet. Unlike say where a cable modem is a cable modem and will work with any CMTS, PON is not there yet. Dave B. On 10/6/2026 9:35 AM, Adam Moffett wrote: > They're technically ITU standards, and usually you can get 3rd party ONT's > passing traffic on your OLT, but you likely lose the management and > monitoring features, so you still want to replace them but it doesn't have to > be an immediate forklift. > > I'd say the most valuable things the provider has are their pole attachments, > conduit leases, and buried conduit. It doesn't really cost a fortune to hang > a cable on a pole, at least not compared to the cost of make-ready, > engineering, and licensing. If you hate their entire infrastructure you can > overlash a new network on top of theirs for a fraction of the original cost. > > There is a lot of variety in design choices, and I would say another ISP is > more valuable if they built their infrastructure similar to how you would do > it, but they probably didn't. > > -Adam > > > ________________________________ > From: AF <[email protected]> on behalf of Ken Hohhof <[email protected]> > Sent: Tuesday, October 6, 2026 9:08 AM > To: 'AnimalFarm Microwave Users Group' <[email protected]> > Subject: Re: [AFMUG] PON and UPC > > > I’ve noticed a lot of medium to large FTTH providers are very heavily > leveraged, all their fiber assets are pledged as collateral to get loans to > build more. I would guess as BEAD rolls out, the copays are also borrowed > money. My gut tells me if revenue fails to meet projections there could be a > lot of loan defaults. > > > > Which gets me to my question. In the popular fiber architectures, what parts > are generic and can easily be used by a new owner? The glass itself would > seem to be useful no matter what. But what about ONTs, OLTs, and customer > prem equipment like routers? And other headend equipment like routers. > > > > What happens if a fiber equipment manufacturer like Calix goes though a > bankruptcy or liquidation (thinking about the Cambium situation). Is this > stuff all proprietary, closed ecosystem? > > > > From: AF <[email protected]> On Behalf Of Adam Moffett > Sent: Tuesday, October 6, 2026 7:55 AM > To: AnimalFarm Microwave Users Group <[email protected]> > Subject: Re: [AFMUG] PON and UPC > > > > I've made the same switch myself. PON is definitely easier. Everyone > (including me) likes Ethernet when they're new at it because it seems more > simple and familiar, but PON is easier to manage. I'd keep Ethernet around > for enterprise stuff, but bulk tier3 services are better off on PON. > > There was also a business case for Ethernet in a rural area where you might > have small clusters of potential customers. You could add a 24 port or 48 > port Ethernet switch a lot more cheaply than adding an OLT, but in the past > decade you could also use Ubiquiti's UFiber in that same scenario. Or use > XGS-PON where you can get long reach by adjusting min and differential range > and minding your link budget. In that case you can look at it as the cost of > adding ports on a central system vs the cost of adding a new > point-of-presence. > > In other words, I think the tech based argument for Ethernet was at least > partially wrong from the start, and the most compelling business case for it > has evaporated. > > I also see Ethernet in open access networks where they put each provider in a > VLAN and each ISP is just providing a router and customer service. I won't go > on that rant today, but that model seems incredibly dumb in my eyes. > > > > -Adam > > > > > > > > On Mon, Oct 5, 2026 at 5:22 PM Nate Burke > <[email protected]<mailto:[email protected]>> wrote: > > I don't have any PON in the field, only Active, but I'm curious why > you're converting your active network to a PON network? Network > Management? Future Scalability with existing fiber? What's driving the > switch. I'm genuinely curious. > > On 10/5/2026 3:53 PM, Jason McKemie wrote: >> I have an active optical network that I would like to convert to PON. >> Unfortunately, the connectors at the cabinet are all UPC. Am I asking >> for a headache if I connect those customers via an APC to UPC cable vs >> re-splicing everything at the cabinet? The lengths from the cabinet to >> the customer locations are all less than a mile. I'll be splicing on a >> new APC connector at the customer locations. >> > > -- > AF mailing list > [email protected]<mailto:[email protected]> > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > > -- AF mailing list [email protected] http://af.afmug.com/mailman/listinfo/af_af.afmug.com
-- AF mailing list [email protected] http://af.afmug.com/mailman/listinfo/af_af.afmug.com
