From plug-discuss-bounces@lists.phxlinux.org Thu Jun 11 07:37:54 2020 Return-Path: X-Original-To: lurker@lists.phxlinux.org Delivered-To: lurker@lists.phxlinux.org Received: from phxlinux.org (localhost [127.0.0.1]) by phxlinux.org (Postfix) with ESMTP id 6168E32A0007; Thu, 11 Jun 2020 07:37:50 -0700 (MST) X-Original-To: plug-discuss@lists.phxlinux.org Delivered-To: plug-discuss@lists.phxlinux.org Received: from mail-ej1-f46.google.com (mail-ej1-f46.google.com [209.85.218.46]) by phxlinux.org (Postfix) with ESMTPS id 9D00732A0004 for ; Thu, 11 Jun 2020 07:37:47 -0700 (MST) Received: by mail-ej1-f46.google.com with SMTP id w16so6156251ejj.5 for ; Thu, 11 Jun 2020 07:37:47 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20161025; h=mime-version:references:in-reply-to:from:date:message-id:subject:to :cc; bh=LkqCthUQUjBWP8KdR7tat6OWh85aE1Fo2h8mqAYoeZk=; b=rPTeN/IUD7E7rW5YVkeWPF8FmP5rcSMuBBMmG808IRHn5zmXc4pjIUvivC5fosZK2D e7W2KgeWe7wAKdYbeZy8ZW00ZLB16Pdq8OwkOliIrV7eVm8V/TLnJLN93WO7znXU8eRL ygyTD35dp3zDFj7h+rzmJHS9aX4T0kGHIWruX4OGWTJImXUjKKgF7rYx+3e1qI5DEnHy fOcKR93710EddlPah9lnile3fgiSYFpPkWD62+1v9nx4TVRlK4aoI6DurtS+8knIM861 72FHhlNjgpPnODxItEmm4YccvPb3cHs7DaRAaacPeItyTZhnxAi6yiIgtb9bXypDlbK+ YtLQ== X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:mime-version:references:in-reply-to:from:date :message-id:subject:to:cc; bh=LkqCthUQUjBWP8KdR7tat6OWh85aE1Fo2h8mqAYoeZk=; b=fLDorbryiv5rLsKIv6oO0/CK1yz2ZGaSGa5CBsLV1wVT6Lq6JvlO6YbEsh5VSiqlE3 8bHvV7VXgIpt2D2zAPtw/guLIdwbGiaF4m3fic95egTCszUL2X86uFeRJFQDghxsj9+f Ghjlud9qkHm+9qWftwwiOJjkOK07D+HCuxUmJeO2j9d6PIVqNarJ6mrc3iV46r3P2Ked vvmcP6lDAry3RXt+27iN+sbjgKF53UdmJDbA1oDHruRN57u71G7T3kNFDwtqVICweQ7p 7RIIwriYXrQIJvoLl6hCkY75VAioaeZRECYDlUgWFqWN6DzYq/ex6ADYHNf+kSWJmuq/ +uRA== X-Gm-Message-State: AOAM5320G2OA5ndxYUBn2MEGhGjqxX8Pa3vXEB27mhkTgVafu+9HKtM+ EjEXsZaLQSfBaZ+VjOh0pSLFb3veoESrEBdpcijwSRJK2/g= X-Google-Smtp-Source: ABdhPJyDdJN1A4lEL+SOJYlTBbPsdhaelJCR2nTQUd6X38iLGN9wqUd0PEaoiz5XLHJOPKKHb+M0U8phq0pm8RK/8gY= X-Received: by 2002:a17:906:5418:: with SMTP id q24mr8497976ejo.266.1591886264968; Thu, 11 Jun 2020 07:37:44 -0700 (PDT) MIME-Version: 1.0 References: <530b3422-2f21-e1a0-66fe-882a374ce787@oscontext.com> In-Reply-To: Date: Thu, 11 Jun 2020 10:37:33 -0400 Message-ID: Subject: Re: PfSense + ubiquity To: Main PLUG discussion list X-BeenThere: plug-discuss@lists.phxlinux.org X-Mailman-Version: 2.1.20 Precedence: list List-Id: Main PLUG discussion list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , From: Thomas Scott via PLUG-discuss Reply-To: Main PLUG discussion list Cc: Thomas Scott Content-Type: multipart/mixed; boundary="===============4271393506790542326==" Errors-To: plug-discuss-bounces@lists.phxlinux.org Sender: "PLUG-discuss" --===============4271393506790542326== Content-Type: multipart/alternative; boundary="00000000000058ab1205a7cfe743" --00000000000058ab1205a7cfe743 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: quoted-printable Well as of the first, I am no longer employed by the aforementioned service provider. The article was interesting - but I actually wrote the author with the following after reading the article - my question is whether the 8-12GB was upload only, or upload/download Was 8 - 12 TB (TeraBYTES) the combined, or just the upload? From reading > the article it was unclear which was being referred to. > > The reason this number caught my eye is that 12 TB on a 35 Mbps > (MegaBITS/second) connection is literally 100% of a regular 30 day month. > If it's upload only - he's using his upload pegged at 35 Mbps for the > entire time he's online. I recognized that number from looking at > utilization graphs for business class customers (which don't have a cap). > I'm not in any position to comment on this officially (especially as a > former employee), but if someone is uploading 8-12 TB/month, on a shared > medium, which DOCSIS over HFC (Hybrid Fiber/Coaxial) definitely is, they > are the single largest user of upstream bandwidth on the node. > > Here's the fun part - I'm a consumer, I get that what "unlimited" means i= s > an arbitrary and oft debated question. Looking at data caps in the mobile > industry - is unlimited truly ever unlimited? At the other hand, I'm a > professional service provider engineer - those nodes are using shared > upstream sources, and he is utilizing 20% of that upstream on a regular > basis. How do you decide to police that? How much is too much, etc, etc. = As > i said, I can't comment on that - but given that shared upstream is only > 150Mbps on a node - and that nodes serve 100s of customers, he's an > outlier, and *the* outlier, as in "you can't lie out any further". > > Granted that's all an assumption off of that being upload only - if it's > shared up/down, then it's still notable, and he's still an outlier. > > Managing these networks isn't easy - but it is getting better, Cox is > aggressively rolling out Fiber deeper into their nodes and splitting them > using newer techniques that will allow coax to deliver much faster down > speeds (10Gbps if I recall correctly) and everyone's favorite punching ba= ck > on cable - upload (up to 1Gbps). Those are contingent on newer DOCSIS > specs, 3.1 -> 4.0 and some magic caled Full Duplex DOCSIS that allow you = to > use all frequences on the RF plant both directions. > I honestly have no idea how to manage a DSLAM/DSL network, are they on a shared medium like DOCSIS RF? If so individual QoS can be implemented and shared much more easier than an HFC design. Speaking to FTTH builds - Cox is still doing those in greenfield builds, as is CLINK, but it's incredibly expensive to do in brownfield (already built) neighborhoods. I live in one of the areas of GA that has FTTH and love it, but I don't relish the digging that had to be done to accomplish it. Not saying it can't be done, but I get why Google Fiber, AT&T, VZ Fios, et. al have held off for so long or have delayed/canceled their future brownfields. It's not easy - granted we'll see how 5G impacts all of this for eyeball networks. But in a forum like this, I don't see 5G for 10TB uploads anytime soon. Some of us aren't exactly eyeballs only :) Not a huge fan of the caps - but it's a nasty cycle to be in - node splits are at least 50K a pop, and when I was with former employer, they were done more often in high Business traffic areas (higher monthly revenue, no bandwidth caps), but I wouldn't be surprised to see those mitigated by the new OCML (https://broadbandlibrary.com/ocml-for-converged-access-networks/) for those builds now. - Thomas Scott | mr.thomas.scott@gmail.com On Wed, Jun 10, 2020 at 8:04 PM Michael Butash via PLUG-discuss < plug-discuss@lists.phxlinux.org> wrote: > I started getting taxed with Cox when I was experimenting with letting > some family and friends vpn into my storage, and everyone started > downloading off me at once. I wasn't graced to get gigablast, so no > unlimited for me. Go figure, but I got annoyed quick with cox telling me= I > was going over my "allocation" and charging me to boot. I moved to CL, n= o > more overages, and far cheaper overall. Service is meh at times, but see > prior comments. I'll take cheap, usually fast, and allow for leeching fo= r > the fam. > > Leeching, backups, all the same. > > -mb > > > On Wed, Jun 10, 2020 at 8:08 AM Stephen Partington via PLUG-discuss < > plug-discuss@lists.phxlinux.org> wrote: > >> I ended up with fiber to the home in my area, and Have used anywhere fro= m >> 2-10 TB a month since long before its availability. Only once did I rece= ive >> a call, I explained I was doing a backup restore to cloud and that was i= t. >> . I have heard nothing else since. >> >> On Wed, Jun 10, 2020 at 7:58 AM Michael Butash via PLUG-discuss < >> plug-discuss@lists.phxlinux.org> wrote: >> >>> Kind of a funny read, made me think of this Cox discussion. As usual, >>> even when you pay for unlimited, it's not really, and if you piss off a >>> random top-talker metric, you get smacked. Actually get what you pay f= or? >>> Nah. >>> >>> >>> https://arstechnica.com/tech-policy/2020/06/cox-slows-internet-speeds-i= n-entire-neighborhoods-to-punish-any-heavy-users/ >>> >>> I don't buy the FUD about the "downgrade the whole neighborhood", unles= s >>> the neighborhood is just overused/saturated as it is, in which case Cox >>> needs to fix it with a node split per normal direction. They won't >>> police/shape a whole neighborhood like that, rather they'd just >>> decommission or lower the bandwidth on the offenders modem usually, ala >>> this guy. >>> >>> May be a bit different if an actual Cox fiber/pon site, , but these see= m >>> still rare like hens teeth, and only was deployed as buzz during Google >>> Fiber threatening them. Cox doing fiber to the home I think died with >>> Google Fiber. >>> >>> -mb >>> >>> >>> On Thu, May 14, 2020 at 9:32 AM Michael Butash >>> wrote: >>> >>>> I'll agree with the CL being saturated comment - pretty sure it doesn'= t >>>> matter whether DSL or Fiber, their peering and aggregation is the same= per >>>> region, and really it's where they converge that is the problem, which= is >>>> where said saturation occurs. CL just *feels* saturated in use, where= I >>>> didn't get that with Cox. Everything loads a little slower, you can j= ust >>>> sort of tell after using long enough. Cox would periodically too, but= they >>>> tended to already be working on a fix by the time I'd hit up someone I= knew >>>> there to complain. CL I have no such faith in. >>>> >>>> I'm paying almost half my Cox bill with CL however, and no random >>>> overage charges, so I'm willing to live with it honestly, and it's nev= er >>>> been *that bad*. If I download something, it downloads quickly, be it= http >>>> or torrents. Just random viewing of pages in quick succession, ala >>>> scanning news just always seems a bit slow to start. That usually fee= ls >>>> like buffers are blown out somewhere inline. >>>> >>>> -mb >>>> >>>> >>>> On Wed, May 13, 2020 at 8:34 PM Thomas Scott via PLUG-discuss < >>>> plug-discuss@lists.phxlinux.org> wrote: >>>> >>>>> They are welcome to, but node splits are a 6 month minimum last I >>>>> checked =F0=9F=98=81 - granted we're getting faster with how many we'= re doing. In the >>>>> next 5 years, most cable operators will implement some sort of >>>>> aggressive node splitting to keep up with demand. Current employer no= t >>>>> excluded. >>>>> >>>>> I've had CLink on fiber - they're upstream nodes are a little more >>>>> saturated, but they do peer locally in the valley. Current employer d= oes >>>>> have peering with FAANG and a couple other heavy hitters in the valle= y (not >>>>> any proprietary information here, any trace route from the valley to = those >>>>> sites will show it terminating in 2 or 3 hops), but if I recall corre= ctly >>>>> 70% of CLink traffic hits their DCs in Phoenix. Granted it's all best >>>>> effort past that, but if you don't have a heavily saturated node, you= 'll do >>>>> all right. GPON fiber is GPON fiber, regardless of Service Provider. = It's >>>>> just a question of how many other subscribers are on your PON port an= d how >>>>> big the upstream links are. >>>>> >>>>> - Thomas Scott | mr.thomas.scott@gmail.com >>>>> >>>>> >>>>> >>>>> On Tue, May 12, 2020 at 4:04 PM Stephen Partington via PLUG-discuss < >>>>> plug-discuss@lists.phxlinux.org> wrote: >>>>> >>>>>> This last bit is interesting. I have Cox Fiber (no data cap for >>>>>> Gigablast fiber yet) and Century Link just announced a competing ser= vice in >>>>>> my area. For about half the cost. For the same Gigabit Fiber (or 940= mbps as >>>>>> they are calling it). >>>>>> >>>>>> Anyone with any experience with them on residential fiber? >>>>>> >>>>>> On Tue, May 12, 2020 at 5:59 AM Michael Butash via PLUG-discuss < >>>>>> plug-discuss@lists.phxlinux.org> wrote: >>>>>> >>>>>>> So Cox subs can reach out to you when we're having saturation >>>>>>> issues? :) >>>>>>> >>>>>>> Having been around for the beginnings of cable modem tech at @home >>>>>>> networks in the 90's dealing with almost every big MSO (Cox, Comcas= t, ATT, >>>>>>> Intermedia, etc), I like to talk about the tech as a bit proud wher= e it's >>>>>>> gone. I liked Cox as one of the last decent hold-outs for things l= ike >>>>>>> keeping Usenet around longer than they should, not killing customer= s for >>>>>>> mpaa/riaa abuse complaints, and keeping data caps off when the indu= stry was >>>>>>> moving in that direction, so I think they're better than the rest, = but >>>>>>> eventually they hopped on the money train with data caps too. And = now >>>>>>> they're paying for their pro-pirate stance as well with lawsuits >>>>>>> against them winning >>>>>>> , >>>>>>> probably using that extra cap revenue to pay the trolls. >>>>>>> >>>>>>> Would I go back? Not as long as they have data caps, and someone >>>>>>> else around me doesn't, but yes - much better network. I don't lik= e random >>>>>>> overages in my bill, I get that enough with power. If I thought th= e covid >>>>>>> restrictions to remove caps would hold, I'd probably switch back no= w, but >>>>>>> I'm sure they'll find a reason to reimplement them asap as that's l= ost >>>>>>> revenue on your rsu's. >>>>>>> >>>>>>> It's always good to hear from other docsis speakers, welcome back! >>>>>>> >>>>>>> -mb >>>>>>> >>>>>>> >>>>>>> On Mon, May 11, 2020 at 6:54 PM Thomas Scott < >>>>>>> mr.thomas.scott@gmail.com> wrote: >>>>>>> >>>>>>>> Day job is for a certain ISP HQ in Atlanta that supplies internet >>>>>>>> for a lot of the valley - I work in Network Operations first in Ph= oenix and >>>>>>>> now in Atlanta, and was surprised to see so much of what I talk >>>>>>>> about everyday in PLUG! >>>>>>>> >>>>>>>> CLink trying to play FTTN as FTTH, nothing new there. I live in a >>>>>>>> neighborhood outside of Atlanta that had some AT&T brownfield deve= lopment >>>>>>>> for FTTH, and I've had no regrets (300 up 300 down!) Cox is moving= towards >>>>>>>> "10G" with DOCSIS 4.0 and they are getting fiber closer to the hom= e with >>>>>>>> their node splits. If you find that you all off a sudden have an e= xtra hop >>>>>>>> in your path, that might be the seen you've been on one of those n= odes that >>>>>>>> have been lit and split. The amount of bandwidth going up and down= will go >>>>>>>> up dramatically. >>>>>>>> >>>>>>>> @Michael - yeah I don't think the caps are going anywhere, the >>>>>>>> industry as a whole (driven by big red) has moved that direction, = but I >>>>>>>> think you'll see speeds and caps rise as N+0 goes to full duplex D= OCSIS. I >>>>>>>> do know they've been relaxed with the COVID-19 FCC initiatives, bu= t how >>>>>>>> long that lasts, I'm not sure. >>>>>>>> >>>>>>>> @Mac - the cox supplied modems are almost all going to "Panoramic >>>>>>>> Wi-Fi" and the number of holes found in DOCSIS devices is... distu= rbing to >>>>>>>> say the least. It was designed to be operated on a shared RF mediu= m, and >>>>>>>> like other "trusting" protocols (i.e. BGP) has a lot of issues. Th= e more >>>>>>>> virtualized it becomes, I think we'll see more of those go away - = the >>>>>>>> smaller the broadcast domains, and the smaller the first upstream = router, >>>>>>>> the better those will be able to be maintained and automated. Look= ing at >>>>>>>> the road maps, it will be interesting what comes next. >>>>>>>> >>>>>>>> - Thomas Scott | mr.thomas.scott@gmail.com >>>>>>>> >>>>>>>> >>>>>>>> >>>>>>>> On Wed, May 6, 2020 at 3:54 PM Michael Butash via PLUG-discuss < >>>>>>>> plug-discuss@lists.phxlinux.org> wrote: >>>>>>>> >>>>>>>>> Oddly enough, the model number of your router stuck in my head, >>>>>>>>> the C3000Z, and I realized I used the same thing, but for my 150m= bps dsl >>>>>>>>> modem. You sure you have actual gig fiber? They tend to misrepr= esent >>>>>>>>> their actual products in sales. Ask me how I know. >>>>>>>>> >>>>>>>>> >>>>>>>>> >>>>>>>>> I say this because I called CL before going to them, and asked if >>>>>>>>> I could get fiber in the network. They said yes. Hmm, I knew da= mn well >>>>>>>>> they did not, as no one wants to build fiber into old peoria neig= hborhoods >>>>>>>>> such as mine. After some conversation and calling him out, he ex= plained >>>>>>>>> that "oh, it's a gigabit network", just not fiber to your house. = I could >>>>>>>>> get dual-band DSL, which means 75mbps x2, for a total of 150mbps,= delivered >>>>>>>>> by a gigabit network! I sort of facepalmed, but ordered it anywa= ys as it >>>>>>>>> was significantly more than I had with cox (80mbps at the time I = think), >>>>>>>>> significantly cheaper, and no bandwidth cap. >>>>>>>>> >>>>>>>>> If there is anything other than fiber directly in your modem, I'd >>>>>>>>> call bullocks, but FTTH is a myth to me. >>>>>>>>> >>>>>>>>> Crappier service, but I'll take the (usually) cheap and fast. It >>>>>>>>> is most certainly not gigabit fiber to my house, even though that= 's what >>>>>>>>> they tried to sell me I was getting. Only new house/community bu= ilds get >>>>>>>>> fiber, and if even that. Cox did the same to compete with Google= fiber, >>>>>>>>> and as soon as Google Fiber died, so did Cox ever mentioning fibe= r again. >>>>>>>>> Truth is Cox doesn't need it, shielded coax can deliver soon 10g = over it >>>>>>>>> just fine with new modulation schemas and docsis improvements. >>>>>>>>> Centurylink's 100 year old 2-8 wire infrastructure cannot, all th= ey can do >>>>>>>>> is build new with fiber, but they probably won't being decrepit. >>>>>>>>> >>>>>>>>> I hear friends of mine mention they have fiber, and wonder just i= f >>>>>>>>> they really do. This is why Google Fiber folded, it was unrealis= tic unless >>>>>>>>> a net-new community build. Google fiber retrofits were a disaste= r >>>>>>>>> >>>>>>>>> . >>>>>>>>> >>>>>>>>> Fun-fact: Oddly enough the guy that built Google Fiber, Milo >>>>>>>>> Medin, is the same guy that started @Home Networks back in late 9= 0's for >>>>>>>>> Cable Modem services, and pioneered current industry standards in= use today >>>>>>>>> globally to deliver cable internet. The last-mile regional MSO p= roviders >>>>>>>>> snuffed him/company back then, took it over themselves, and then = they >>>>>>>>> snuffed him out again as he tried the same incursion with Google = Fiber, and >>>>>>>>> realized it just cost too damn much to compete. Cable Monopolies= , flawless >>>>>>>>> victory. >>>>>>>>> >>>>>>>>> Next I expect he'll team up with Elon or Bezos to try again via >>>>>>>>> terrestrial. >>>>>>>>> >>>>>>>>> -mb >>>>>>>>> >>>>>>>>> >>>>>>>>> On Wed, May 6, 2020 at 10:32 AM Michael Butash >>>>>>>>> wrote: >>>>>>>>> >>>>>>>>>> I tend to find the CL network a bit wonky, having moved to DSL >>>>>>>>>> from Cox (damn bandwidth caps). I find the general performance = is worse >>>>>>>>>> than cox, where I suspect they simply don't manage the bandwidth= and are >>>>>>>>>> far too oversubscribed as it feels like the internet buffers at = times, >>>>>>>>>> literally. Cox would occasionally get that way too, and it was = easy to see >>>>>>>>>> in an ongoing MTR when their peering in LA would get slammed and= latency >>>>>>>>>> would jump (not to mention I know the guys that manage that band= width, >>>>>>>>>> telling them often got it fixed). Oddly Using MTR with CL, they= filter >>>>>>>>>> icmp/udp specifically that seems to hide responses to track well= . Go >>>>>>>>>> figure, truth hurts, so hide it. >>>>>>>>>> >>>>>>>>>> Having worked for service providers numerous times over the >>>>>>>>>> years, working in and building them, routers are always an issue= in a metro >>>>>>>>>> city or even interstate networks. No two platforms are ever the= same, >>>>>>>>>> whether buying all Cisco, Juniper, Nokia, or any combo of all an= d more, >>>>>>>>>> which as you said, many do. Hardest part is usually capacity pl= anning, >>>>>>>>>> particularly with something like covid, every isp took a kick in= the groin >>>>>>>>>> at the same time to augment their networks, suddenly by some mag= nitude, >>>>>>>>>> when everyone else in the world is doing the same. Slowness in = networking >>>>>>>>>> can often be attributed to those not having enough capacity, tho= ugh they'll >>>>>>>>>> never admit it. >>>>>>>>>> >>>>>>>>>> I'm on the 150mbps dsl, and a speed test can provide that for >>>>>>>>>> sure, but general usage, which I use a lot of tabs and apps, ten= ds to bring >>>>>>>>>> things to a crawl often. I'd even go back to cox if they got ri= d of the >>>>>>>>>> bandwidth cap. CL might as well be government, and they're run = by unions, >>>>>>>>>> so nothing happens fast, including capacity augments. >>>>>>>>>> >>>>>>>>>> Re: mac limits, having been around Cox both as a customer and >>>>>>>>>> network engineer working there early 2000's, the mac security wa= s more >>>>>>>>>> about limiting the amount of hosts behind a modem that could be = allowed to >>>>>>>>>> a single mac and IP address. Back Circa 1998 I had my first Co= x modem, >>>>>>>>>> and there were no routers, you just got yourself a phat 10baset = switch from >>>>>>>>>> computer city and connected up your family on public ip addresse= s, each >>>>>>>>>> with their own mac and ip's. With no limits or filters that led= to >>>>>>>>>> security issues (hey, I see my neighbor's c drive shared!), Cox = and others >>>>>>>>>> then pushed people to then buy a router, which by then around 20= 02, you >>>>>>>>>> could buy a cheap wrt54g linksys. The advent of docsis also all= owed to >>>>>>>>>> both filter and restrict the macs by default, also let them redu= ce to now >>>>>>>>>> 1:1 IP to User ratio, which was good for ip management, the abus= e >>>>>>>>>> departments, and fbi warrants from legal. You used to be able t= o buy >>>>>>>>>> another ip, they'd push a new docsis config with mac-alowed=3D2,= but not >>>>>>>>>> anymore. >>>>>>>>>> >>>>>>>>>> Same reasons they're just building in the router functions now, >>>>>>>>>> it ensures they can offer some basic customer security, plus let= s them run >>>>>>>>>> whatever spyware in their embedded router os they want. Better = off buying >>>>>>>>>> your own standalone modem and router combo, one you ideally trus= t. >>>>>>>>>> >>>>>>>>>> -mb >>>>>>>>>> >>>>>>>>>> >>>>>>>>>> On Tue, May 5, 2020 at 10:07 PM Donald Mac McCarthy via >>>>>>>>>> PLUG-discuss wrote: >>>>>>>>>> >>>>>>>>>>> Putting a CL modem into a bridge mode where it only handles the >>>>>>>>>>> PPPoE connection is simply checking a radial select button and = hitting >>>>>>>>>>> apply. If your firewall supports PPoE, even better, as you no l= onger need >>>>>>>>>>> their Modem and router in the mix. But, that is just my experie= nce, and it >>>>>>>>>>> is limited. I have a CL fiber to the door drop, and they gave m= e a Zyxel >>>>>>>>>>> C3000Z device for connection. I promptly ripped it out and allo= wed pfSense >>>>>>>>>>> to maintain the PPPoE connection. I had to call support for pac= ket loss one >>>>>>>>>>> time, and they refused to help me. So goes it rolling your own = I guess. >>>>>>>>>>> Turns out a day later we had a several hour outage due to one o= f the >>>>>>>>>>> multiplexing cards used to distribute the 40Gb/s core fiber to = the GPON >>>>>>>>>>> devices failed. Seems like that was a likely culprit for some o= f the packet >>>>>>>>>>> loss the previous day. >>>>>>>>>>> >>>>>>>>>>> Having just gotten off a call in which the Senior Director of >>>>>>>>>>> Security Architecture and Engineering (a friend of mine from At= lanta) for >>>>>>>>>>> Cox was a participant, before he hung up I asked him about the = typical Cox >>>>>>>>>>> supplied modems. Very, very few of them are purely bridge devic= es - >>>>>>>>>>> especially with the push to "Panoramic WiFi". A member of Centr= yLink who >>>>>>>>>>> was also on the call (ISP InfoSec sharing/working group) mentio= ned how >>>>>>>>>>> painful it was to support the number of company issued >>>>>>>>>>> modems/gateway/router models there are for different infrastruc= ture and >>>>>>>>>>> connections - let alone ones that customers buy and bring to th= e party. >>>>>>>>>>> BTW, the MAC address thing is because they do actually use a MA= C locking >>>>>>>>>>> like feature for security. Apparently it is bad for the network= if you just >>>>>>>>>>> go plug your modem in at several houses in the neighborhood due= to the way >>>>>>>>>>> DOCSIS works. I still have to dig into that and ask some more q= uestions on >>>>>>>>>>> that one. >>>>>>>>>>> >>>>>>>>>>> There was a collective groan among the engineers when another >>>>>>>>>>> ISP spoke up about the number of critical flaws they find in th= eir DOCIS >>>>>>>>>>> devices each year. >>>>>>>>>>> >>>>>>>>>>> With the amount of consolidation which has happened in the past >>>>>>>>>>> 20 years in the broadband market, the landscape is riddled with= legacy bits >>>>>>>>>>> and pieces of this provider and that provider somehow being coe= rced into >>>>>>>>>>> working together to accomplish passing traffic. One of the ISPs= mentioned >>>>>>>>>>> they had no less than 350 different models of core switching eq= uipment made >>>>>>>>>>> by more than a dozen manufacturers in their network. They have = a team of 40 >>>>>>>>>>> (really 5 teams of 8) that simply monitor and ensure that the O= SPF >>>>>>>>>>> functions properly among the various models and brands to make = sure that >>>>>>>>>>> the network properly heals/manages congestion. >>>>>>>>>>> >>>>>>>>>>> Anyway, just throwing it out so that people can see and >>>>>>>>>>> understand the picture at a higher level. The final comment on = the call was >>>>>>>>>>> from an engineer at a midwestern rural provider and one that I = am sure many >>>>>>>>>>> of us can relate to. She said she spends all day pulling her ha= ir out >>>>>>>>>>> trying to keep the network functioning at the highest of levels= . The first >>>>>>>>>>> words out of her kids' mouths when she gets home are "Mom, the = WiFi seems >>>>>>>>>>> slow today." >>>>>>>>>>> >>>>>>>>>>> I talked with Alexander this afternoon, and it looks like he ha= s >>>>>>>>>>> a functioning network again. The APs were reluctant to give up = their old >>>>>>>>>>> configuration, so a factory reset and new DHCP leases seem to h= ave done the >>>>>>>>>>> trick. >>>>>>>>>>> >>>>>>>>>>> Hopefully this sheds a bit of light on something for a few >>>>>>>>>>> people. >>>>>>>>>>> >>>>>>>>>>> Mac >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> Michael Butash via PLUG-discuss wrote on 5/4/20 4:59 PM: >>>>>>>>>>> >>>>>>>>>>> Ideally when you plug into a cable modem, it comes up, and >>>>>>>>>>> passes your ethernet to the cmts in a bridge, lets one mac addr= ess >>>>>>>>>>> dhcp/arp, and things work. It learns that one ip/mac, and disa= llows any >>>>>>>>>>> other mac. No security, nat, nothing, just real dumb dhcp + de= fault >>>>>>>>>>> routing with a public ip. Routers/firewalls try to NAT you, th= us double >>>>>>>>>>> NAT if using a router behind it. >>>>>>>>>>> >>>>>>>>>>> CL sells you a dsl modem/router that does your local security >>>>>>>>>>> whether you want it or not, full router/nat/firewall, and proba= bly >>>>>>>>>>> spyware. Making it a modem is possible, but takes work, and yo= ur firewall >>>>>>>>>>> has to support PPPoE (not all can/do). Last time I touched a c= ombo Cox >>>>>>>>>>> router/modem, I didn't see any way to do so. I told them to bu= y a real >>>>>>>>>>> modem, and that worked with their belkin/cisco/linksys/netgear = they had. >>>>>>>>>>> >>>>>>>>>>> If your "modem" mentions wifi, it's a router/firewall, not a >>>>>>>>>>> modem. Not all are clear about this, as they dumb it down for = consumers, >>>>>>>>>>> but an important point. >>>>>>>>>>> >>>>>>>>>>> -mb >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> On Mon, May 4, 2020 at 1:53 PM Stephen Partington via >>>>>>>>>>> PLUG-discuss wrote: >>>>>>>>>>> >>>>>>>>>>>> I Owned a Nighthawk Router/Modem combo, The way that Netgear >>>>>>>>>>>> handled that is that the modem was hard-wired to a bridge on t= he router >>>>>>>>>>>> side. and technically you could see it as a separate device in= the router >>>>>>>>>>>> configs if you rooted around enough. but the modem side was ju= st a modem. >>>>>>>>>>>> >>>>>>>>>>>> On Mon, May 4, 2020 at 11:03 AM Michael Butash via PLUG-discus= s >>>>>>>>>>>> wrote: >>>>>>>>>>>> >>>>>>>>>>>>> Cox modems *are* bridges first and foremost typically, unless >>>>>>>>>>>>> you get a bundled router/modem, which is only what CenturyLin= k sells. If >>>>>>>>>>>>> you got a "router/modem" combo, just buy a modem-only device = for a dumb >>>>>>>>>>>>> bridge and simple ethernet for a public ip. I recommend stay= ing with an >>>>>>>>>>>>> arris cable modem, originally motorola, they basically develo= ped cable >>>>>>>>>>>>> modem docsis, and are always the best. >>>>>>>>>>>>> >>>>>>>>>>>>> I moved from Cox to CL when Cox started adding a usage cap, >>>>>>>>>>>>> and that was new to me to get my Fortinet firewall online wit= h CL and their >>>>>>>>>>>>> DSL doing PPPOE. I've seen the router/cable modem combo boxe= s later, but >>>>>>>>>>>>> never owned one as I always have my own router/firewall. >>>>>>>>>>>>> >>>>>>>>>>>>> -mb >>>>>>>>>>>>> >>>>>>>>>>>>> >>>>>>>>>>>>> On Mon, May 4, 2020 at 8:36 AM Donald Mac McCarthy < >>>>>>>>>>>>> mac@oscontext.com> wrote: >>>>>>>>>>>>> >>>>>>>>>>>>>> Will Cox allow for a bridge/virtual bridge mode? Xfinity >>>>>>>>>>>>>> does, which allows you to put in a firewall, and use the mod= em only as a >>>>>>>>>>>>>> gateway, therefore preventing a double NAT situation. Never = lived in a Cox >>>>>>>>>>>>>> area before, and currently ride CL fiber. >>>>>>>>>>>>>> >>>>>>>>>>>>>> Mac >>>>>>>>>>>>>> >>>>>>>>>>>>>> Michael Butash via PLUG-discuss wrote on 5/3/20 2:00 PM: >>>>>>>>>>>>>> >>>>>>>>>>>>>> Cox modems will learn and allow only 1 mac at a time (unless >>>>>>>>>>>>>> business is set to allow more, but not on residential). If = switching out >>>>>>>>>>>>>> firewalls, I 99% of time reboot the modem first and foremost= . >>>>>>>>>>>>>> >>>>>>>>>>>>>> -mb >>>>>>>>>>>>>> >>>>>>>>>>>>>> On Sun, May 3, 2020 at 12:08 PM Snyder, Alexander J via >>>>>>>>>>>>>> PLUG-discuss wrote: >>>>>>>>>>>>>> >>>>>>>>>>>>>>> I got it working. >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> I assigned the SFP+ port as my LAN and assigned it the >>>>>>>>>>>>>>> 10.x.x.x/16 network. Then I had to call COX and list the WA= N Mac address >>>>>>>>>>>>>>> with them. Upon doing so I was able to reach external sites= , and all >>>>>>>>>>>>>>> downstream devices started coming alive! >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Thanks for all the suggestions and help! >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Thanks, >>>>>>>>>>>>>>> Alexander >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> Sent from my Galaxy S10+ >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> On Sun, May 3, 2020, 03:14 Herminio Hernandez, Jr. via >>>>>>>>>>>>>>> PLUG-discuss wrote: >>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> Can you login to the FW via the LAN interface? Can you >>>>>>>>>>>>>>>> ping the FW LAN interface? Check the routing and NAT polic= y on the FW. All >>>>>>>>>>>>>>>> outbound traffic should NAT to the FW WAN interface and th= ere should be a >>>>>>>>>>>>>>>> default (0.0.0.0/0) route to the internet. >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> On Sat, May 2, 2020 at 7:27 PM Seabass via PLUG-discuss < >>>>>>>>>>>>>>>> plug-discuss@lists.phxlinux.org> wrote: >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> I'm with Mac, I think it is not the firewall, but if you >>>>>>>>>>>>>>>>> have the ability to plug it into a display with a keyboar= d, you can use >>>>>>>>>>>>>>>>> that for configuration and modify a different device at t= he same time. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Makes it easier to troubleshoot by giving you the ability >>>>>>>>>>>>>>>>> to configure your pfSense ports at the same time. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Message: 2 >>>>>>>>>>>>>>>>> Date: Sat, 2 May 2020 09:04:35 -0700 >>>>>>>>>>>>>>>>> From: Donald Mac McCarthy >>>>>>>>>>>>>>>>> To: "Snyder, Alexander J via PLUG-discuss" >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Subject: Re: pfSense + Ubiquity >>>>>>>>>>>>>>>>> Message-ID: < >>>>>>>>>>>>>>>>> 18adfa38-3e72-7b0a-e31a-1ddf175d717f@oscontext.com> >>>>>>>>>>>>>>>>> Content-Type: text/plain; charset=3D"utf-8" >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> I can help - but I am unavailable to do so until tomorrow= . >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Make sure there are not any thing other than default VLAN= s >>>>>>>>>>>>>>>>> on the >>>>>>>>>>>>>>>>> interfaces to start with. Ubiquiti is famous for not >>>>>>>>>>>>>>>>> havinght eSFP+ >>>>>>>>>>>>>>>>> ports active in the default configuration, and I believe >>>>>>>>>>>>>>>>> the switch has >>>>>>>>>>>>>>>>> all the ports to shutdown on default config as well. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> I think it is the switch not passing traffic through - no >>>>>>>>>>>>>>>>> the firewall. >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Mac >>>>>>>>>>>>>>>>> Snyder, Alexander J via PLUG-discuss wrote on 5/2/20 8:53 >>>>>>>>>>>>>>>>> AM: >>>>>>>>>>>>>>>>> > Does anyone out there have experience with pfSence and >>>>>>>>>>>>>>>>> Ubiquity switches? >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > I have zero with either but that didn't stop me from >>>>>>>>>>>>>>>>> buying both .... >>>>>>>>>>>>>>>>> > how hard could it be?! LOL. >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > I bought a Negate XG-1537-1U. I bought a Unifi Pro 24 >>>>>>>>>>>>>>>>> PoE switch. >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > I can configure the FW immediately after >>>>>>>>>>>>>>>>> > firstboot/restore-default-configs, but only if i set th= e >>>>>>>>>>>>>>>>> LAN interface >>>>>>>>>>>>>>>>> > to be the cable that goes directly to my laptop. That's >>>>>>>>>>>>>>>>> great, but >>>>>>>>>>>>>>>>> > that does shit for the downstream switch. >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > I have a 10GB SFP+ Port that I want to configure as the >>>>>>>>>>>>>>>>> downstream >>>>>>>>>>>>>>>>> > port to ubiquity, but any configuration other than >>>>>>>>>>>>>>>>> mentioned above >>>>>>>>>>>>>>>>> > fails .... and I'm now on my 12th "Reset To Factory >>>>>>>>>>>>>>>>> Defaults" ... any >>>>>>>>>>>>>>>>> > help on this would be greatly appreciated! >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > Thanks, >>>>>>>>>>>>>>>>> > Alexander >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > Sent from my Galaxy S10+ >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> > --------------------------------------------------- >>>>>>>>>>>>>>>>> > PLUG-discuss mailing list - >>>>>>>>>>>>>>>>> PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>>>>>> > To subscribe, unsubscribe, or to change your mail >>>>>>>>>>>>>>>>> settings: >>>>>>>>>>>>>>>>> > https://lists.phxlinux.org/mailman/listinfo/plug-discus= s >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> -- >>>>>>>>>>>>>>>>> Donald "Mac" McCarthy >>>>>>>>>>>>>>>>> Director, Field Operations >>>>>>>>>>>>>>>>> Open Source Context >>>>>>>>>>>>>>>>> +1.602.584.4445 >>>>>>>>>>>>>>>>> mac@oscontext.com >>>>>>>>>>>>>>>>> https://oscontext.com >>>>>>>>>>>>>>>>> -------------- next part -------------- >>>>>>>>>>>>>>>>> An HTML attachment was scrubbed... >>>>>>>>>>>>>>>>> URL: < >>>>>>>>>>>>>>>>> http://lists.phxlinux.org/pipermail/plug-discuss/attachme= nts/20200502/aeab14b4/attachment-0001.html >>>>>>>>>>>>>>>>> > >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> ------------------------------ >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> Subject: Digest Footer >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> _______________________________________________ >>>>>>>>>>>>>>>>> PLUG-discuss mailing list - >>>>>>>>>>>>>>>>> PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail setting= s: >>>>>>>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> ------------------------------ >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> End of PLUG-discuss Digest, Vol 179, Issue 2 >>>>>>>>>>>>>>>>> ******************************************** >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>>>>>>> PLUG-discuss mailing list - >>>>>>>>>>>>>>>>> PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail setting= s: >>>>>>>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>>>>>> >>>>>>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.or= g >>>>>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings= : >>>>>>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>>>>> >>>>>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings:h= ttps://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>>>> >>>>>>>>>>>>>> >>>>>>>>>>>>>> -- >>>>>>>>>>>>>> Donald "Mac" McCarthy >>>>>>>>>>>>>> Director, Field Operations >>>>>>>>>>>>>> Open Source Context >>>>>>>>>>>>>> +1.602.584.4445 >>>>>>>>>>>>>> mac@oscontext.com >>>>>>>>>>>>>> https://oscontext.com >>>>>>>>>>>>>> >>>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> >>>>>>>>>>>> -- >>>>>>>>>>>> A mouse trap, placed on top of your alarm clock, will prevent >>>>>>>>>>>> you from rolling over and going back to sleep after you hit th= e snooze >>>>>>>>>>>> button. >>>>>>>>>>>> >>>>>>>>>>>> Stephen >>>>>>>>>>>> >>>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings:http= s://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>>> >>>>>>>>>>> >>>>>>>>>>> -- >>>>>>>>>>> Donald "Mac" McCarthy >>>>>>>>>>> Director, Field Operations >>>>>>>>>>> Open Source Context >>>>>>>>>>> +1.602.584.4445 >>>>>>>>>>> mac@oscontext.com >>>>>>>>>>> https://oscontext.com >>>>>>>>>>> --------------------------------------------------- >>>>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>>>> >>>>>>>>>> --------------------------------------------------- >>>>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>>>> >>>>>>>> --------------------------------------------------- >>>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>>> >>>>>> >>>>>> >>>>>> -- >>>>>> A mouse trap, placed on top of your alarm clock, will prevent you >>>>>> from rolling over and going back to sleep after you hit the snooze b= utton. >>>>>> >>>>>> Stephen >>>>>> >>>>>> --------------------------------------------------- >>>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>>> >>>>> --------------------------------------------------- >>>>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>>>> To subscribe, unsubscribe, or to change your mail settings: >>>>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >>>> >>>> --------------------------------------------------- >>> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >>> To subscribe, unsubscribe, or to change your mail settings: >>> https://lists.phxlinux.org/mailman/listinfo/plug-discuss >> >> >> >> -- >> A mouse trap, placed on top of your alarm clock, will prevent you from >> rolling over and going back to sleep after you hit the snooze button. >> >> Stephen >> >> --------------------------------------------------- >> PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org >> To subscribe, unsubscribe, or to change your mail settings: >> https://lists.phxlinux.org/mailman/listinfo/plug-discuss > > --------------------------------------------------- > PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org > To subscribe, unsubscribe, or to change your mail settings: > https://lists.phxlinux.org/mailman/listinfo/plug-discuss --00000000000058ab1205a7cfe743 Content-Type: text/html; charset="UTF-8" Content-Transfer-Encoding: quoted-printable
Well as of the first, I am no longer employed by the afo= rementioned service provider. The article was interesting - but I actually = wrote the author with the following after reading the article - my question= is whether the 8-12GB was upload only, or upload/download

Was 8= - 12 TB (TeraBYTES) the combined, or just the upload? From reading the art= icle it was unclear which was being referred to.=C2=A0

The reason this number= caught my eye is that 12 TB on a 35 Mbps (MegaBITS/second) connection is l= iterally 100% of a regular 30 day month. If it's upload only - he's= using his upload pegged at 35 Mbps for the entire time he's online. I = recognized that number from looking at utilization graphs for business clas= s customers (which don't have a cap). I'm not in any position to co= mment on this officially (especially as a former employee), but if someone = is uploading 8-12 TB/month, on a shared medium, which DOCSIS over HFC (Hybr= id Fiber/Coaxial) definitely=C2=A0is, they are the single largest user of u= pstream bandwidth on the node.=C2=A0

=
Here's the fun part - I'm a cons= umer, I get that what "unlimited" means is an arbitrary and oft d= ebated question. Looking at data caps in the mobile industry - is unlimited= truly ever unlimited? At the other hand, I'm a professional service pr= ovider engineer - those nodes are using shared upstream sources, and he is = utilizing 20% of that upstream on a regular basis. How do you decide to pol= ice that? How much is too much, etc, etc. As i said, I can't comment on= that - but given that shared upstream is only 150Mbps on a node - and that= nodes serve 100s of customers, he's an outlier, and=C2=A0the=C2= =A0outlier, as in "you can't lie out any further".=C2=A0

Grante= d that's all an assumption off of that being upload only - if it's = shared up/down, then it's still notable, and he's still an outlier.= =C2=A0

Managing these networks isn't easy - but it is getting better, Cox= is aggressively rolling out Fiber deeper into their nodes and splitting th= em using newer techniques that will allow coax to deliver much faster down = speeds (10Gbps if I recall correctly) and everyone's favorite punching = back on cable - upload (up to 1Gbps). Those are contingent on newer DOCSIS = specs, 3.1 -> 4.0 and some magic caled=C2=A0Full Duplex DOCSIS that allo= w you to use all frequences=C2=A0on the RF plant both directions.=C2=A0

I honestly have no idea how to manage a = DSLAM/DSL network, are they on a shared medium like DOCSIS RF? If so indivi= dual QoS can be implemented and shared much more easier than an HFC design.= =C2=A0

Speaking to FTTH builds - Cox is still doin= g those in greenfield builds, as is CLINK, but it's incredibly expensiv= e to do in brownfield (already built) neighborhoods. I live in one of the a= reas of GA that has FTTH and love it, but I don't relish the digging th= at had to be done to accomplish it. Not saying it can't be done, but I = get why Google Fiber, AT&T, VZ Fios, et. al have held off for so long o= r have delayed/canceled their future brownfields. It's not easy - grant= ed we'll see how 5G impacts all of this for eyeball networks. But in a = forum like this, I don't see 5G for 10TB uploads anytime soon. Some of = us aren't exactly eyeballs only :)

Not a h= uge fan of the caps - but it's a nasty cycle to be in - node splits are= at least 50K a pop, and when I was with former employer, they were done mo= re often in high Business traffic areas (higher monthly revenue, no bandwid= th caps), but I wouldn't be surprised to see those mitigated by the new= OCML (https://broadbandlibrary.com/ocml-for-converged-access-networks/) for those builds now.=C2=A0
<= br>

On Wed, Jun 10, 2020 at 8:04 PM Michael Butash via PLUG-discuss <plug-discu= ss@lists.phxlinux.org> wrote:
I started getti= ng taxed with Cox when I was experimenting with letting some family and fri= ends vpn into my storage, and everyone started downloading off me at once.= =C2=A0 I wasn't graced to get gigablast, so no unlimited for me.=C2=A0 = Go figure, but I got annoyed quick with cox telling me I was going over my = "allocation" and charging me to boot.=C2=A0 I moved to CL, no mor= e overages, and far cheaper overall.=C2=A0 Service is meh at times, but see= prior comments.=C2=A0 I'll take cheap, usually fast, and allow for lee= ching for the fam.

Leeching, backups, all the same= .

-mb


On Wed, Jun 10, 2020= at 8:08 AM Stephen Partington via PLUG-discuss <plug-discuss@lists.phxlinux.o= rg> wrote:
I ended up with fiber to the home in my area, and Have used anywhere= from 2-10 TB a month since long before its availability. Only once did I r= eceive a call, I explained I was doing a backup restore to cloud and that w= as it. . I have heard nothing else since.

On Wed, Jun 10, 2020 at 7:58= AM Michael Butash via PLUG-discuss <plug-discuss@lists.phxlinux.org> w= rote:
Kind of a funny read, made me think of this Cox discussion.= =C2=A0 As usual, even when you pay for unlimited, it's not really, and = if you piss off a random top-talker metric, you get smacked.=C2=A0 Actually= get what you pay for?=C2=A0 Nah.


I don't= buy the FUD about the "downgrade the whole neighborhood", unless= the neighborhood is just overused/saturated as it is, in which case Cox ne= eds to fix it with a node split per normal direction.=C2=A0 They won't = police/shape a whole neighborhood like that, rather they'd just decommi= ssion or lower the bandwidth on the offenders modem usually, ala this guy.= =C2=A0

May be a bit different if an actual Co= x fiber/pon site, , but these seem still rare like hens teeth, and only was= deployed as buzz during Google Fiber threatening them.=C2=A0 Cox doing fib= er to the home I think died with Google Fiber.

-mb


On Thu, May 14, 2020 at 9:32 AM Michael Butash <= ;michael@butash.net= > wrote:
=
I'll agree with the CL being saturated comment - = pretty sure it doesn't matter whether DSL or Fiber, their peering and a= ggregation is the same per region, and really it's where they converge = that is the problem, which is where said saturation occurs.=C2=A0 CL just *= feels* saturated in use, where I didn't get that with Cox.=C2=A0 Everyt= hing loads a little slower, you can just sort of tell after using long enou= gh.=C2=A0 Cox would periodically too, but they tended to already be working= on a fix by the time I'd hit up someone I knew there to complain.=C2= =A0 CL I have no such faith in.

I'm paying alm= ost half my Cox bill with CL however, and no random overage charges, so I&#= 39;m willing to live with it honestly, and it's never been *that bad*.= =C2=A0 If I download something, it downloads quickly, be it http or torrent= s.=C2=A0 Just random viewing of pages in quick succession, ala scanning new= s just always seems a bit slow to start.=C2=A0 That usually feels like buff= ers are blown out somewhere inline.

-mb
<= div>

On Wed, May 13, 2020 at 8:34 PM Thomas Scott via PLUG-discu= ss <plug-discuss@lists.phxlinux.org> wrote:
They are welcome to, but node splits are a 6 mo= nth minimum last I checked =F0=9F=98=81 - granted we're getting faster = with how many we're doing. In the next 5 years, most cable operators wi= ll implement some sort of aggressive=C2=A0node splitting=C2=A0to keep up wi= th demand. Current employer not excluded.=C2=A0

I've had CLink on fiber - they're upstream nodes= are a little more saturated, but they do peer locally in the valley. Curre= nt employer does have peering with FAANG and a couple other heavy hitters i= n the valley (not any proprietary=C2=A0information here, any trace route fr= om the valley to those sites will show it terminating in 2 or 3 hops), but = if I recall correctly 70% of CLink traffic hits their DCs in Phoenix. Grant= ed it's all best effort past that, but if you don't have a heavily = saturated node, you'll do all right. GPON fiber is GPON fiber, regardle= ss of Service Provider. It's just a question of how many other subscrib= ers are on your PON port and how big the upstream links are.


On Tue, May 12, 2020 at 4:04 PM Stephen Partington via PLUG-discuss = <pl= ug-discuss@lists.phxlinux.org> wrote:
This last bit is interesting. I have Cox = Fiber (no data cap for Gigablast fiber yet) and Century Link just announced= =C2=A0a competing service in my area. For about half the cost. For the same= Gigabit Fiber (or 940mbps as they are calling it).

Anyone with any experience with = them on residential fiber?

On Tue, May 12, 2020 at 5:59 AM Michael But= ash via PLUG-discuss <plug-discuss@lists.phxlinux.org> wrote:
=
So = Cox subs can reach out to you when we're having saturation issues?=C2= =A0 :)

Having been around for the beginnings of ca= ble modem tech at @home networks in the 90's dealing with almost every = big MSO (Cox, Comcast, ATT, Intermedia, etc), I like to talk about the tech= as a bit proud where it's gone.=C2=A0 I liked Cox as one of the last d= ecent hold-outs for things like keeping Usenet around longer than they shou= ld, not killing customers for mpaa/riaa abuse complaints, and keeping data = caps off when the industry was moving in that direction, so I think they= 9;re better than the rest, but eventually they hopped on the money train wi= th data caps too.=C2=A0 And now they're paying for their pro-pirate sta= nce as well with lawsuits against them winning, probably using that ext= ra cap revenue to pay the trolls.

Would I go back?= =C2=A0 Not as long as they have data caps, and someone else around me doesn= 't, but yes - much better network.=C2=A0 I don't like random overag= es in my bill, I get that enough with power.=C2=A0 If I thought the covid r= estrictions to remove caps would hold, I'd probably switch back now, bu= t I'm sure they'll find a reason to reimplement them asap as that&#= 39;s lost revenue on your rsu's.

It's alwa= ys good to hear from other docsis speakers, welcome back!
-mb


On Mon, May 11, 2020 at 6:54 PM Thomas = Scott <mr= .thomas.scott@gmail.com> wrote:
Day job is for a certain ISP HQ in Atlanta that supplies = internet for a lot of the valley - I work in Network Operations first in Ph= oenix and now in Atlanta, and was surprised to see so much of what I talk a= bout=C2=A0everyday in PLUG!=C2=A0

CLink trying to play FTTN as FTTH, nothing new there. I live in a neig= hborhood outside of Atlanta that had some AT&T brownfield development f= or FTTH, and I've had no regrets (300 up 300 down!) Cox is moving towar= ds "10G" with DOCSIS 4.0 and they are getting fiber closer to the= home with their node splits. If you find that you all off a sudden have an= extra hop in your path, that might be the seen you've been on one of t= hose nodes that have been lit and split. The amount of bandwidth going up a= nd down will go up dramatically.

@Michael - yeah I don't think the caps are going anywhere, the indu= stry as a whole (driven by big red) has moved that direction, but I think y= ou'll see speeds and caps rise as N+0 goes to full duplex DOCSIS. I do = know they've been relaxed with the COVID-19 FCC initiatives, but how lo= ng that lasts, I'm not sure.=C2=A0

@Mac - the cox supplied modems are almost all going to "Pano= ramic Wi-Fi" and the number of holes found in DOCSIS devices is... dis= turbing to say the least. It was designed to be operated on a shared RF med= ium, and like other "trusting" protocols (i.e. BGP) has a lot of = issues. The more virtualized it becomes, I think we'll see more of thos= e go away - the smaller the broadcast domains, and the smaller the first up= stream router, the better those will be able to be maintained and automated= . Looking at the road maps, it will be interesting what comes next.=C2=A0

- Thomas Scott | mr.thomas.scott@gma= il.com=C2=A0


On Wed, May 6, 2020 at 3:54 PM Michael Butash= via PLUG-discuss <plug-discuss@lists.phxlinux.org> wrote:
Oddly enough, the model number of your router stuck in my head, = the C3000Z, and I realized I used the same thing, but for my 150mbps dsl mo= dem.=C2=A0 You sure you have actual gig fiber?=C2=A0 They tend to misrepres= ent their actual products in sales.=C2=A0 Ask me how I know.

=
<tldr>

I say this because I c= alled CL before going to them, and asked if I could get fiber in the networ= k.=C2=A0 They said yes.=C2=A0 Hmm, I knew damn well they did not, as no one= wants to build fiber into old peoria neighborhoods such as mine.=C2=A0 Aft= er some conversation and calling him out, he explained that "oh, it= 9;s a gigabit network", just not fiber to your house.=C2=A0 I could ge= t dual-band DSL, which means 75mbps x2, for a total of 150mbps, delivered b= y a gigabit network!=C2=A0 I sort of facepalmed, but ordered it anyways as = it was significantly more than I had with cox (80mbps at the time I think),= significantly cheaper, and no bandwidth cap.

If t= here is anything other than fiber directly in your modem, I'd call bull= ocks, but FTTH is a myth to me.

Crappier servi= ce, but I'll take the (usually) cheap and fast.=C2=A0 It is most certai= nly not gigabit fiber to my house, even though that's what they tried t= o sell me I was getting.=C2=A0 Only new house/community builds get fiber, a= nd if even that.=C2=A0 Cox did the same to compete with Google fiber, and a= s soon as Google Fiber died, so did Cox ever mentioning fiber again.=C2=A0 = Truth is Cox doesn't need it, shielded coax can deliver soon 10g over i= t just fine with new modulation schemas and docsis improvements.=C2=A0 Cent= urylink's 100 year old 2-8 wire infrastructure cannot, all they can do = is build new with fiber, but they probably won't being decrepit.
<= div>
I hear friends of mine mention they have fiber, and wond= er just if they really do.=C2=A0 This is why Google Fiber folded, it was un= realistic unless a net-new community build.=C2=A0 Google fiber retrofits were a disaster.
=
Fun-fact:=C2=A0 Oddly enough the guy that built Google Fiber= , Milo Medin, is the same guy that started @Home Networks back in late 90&#= 39;s for Cable Modem services, and pioneered current industry standards in = use today globally to deliver cable internet.=C2=A0 The last-mile regional = MSO providers snuffed him/company back then, took it over themselves, and t= hen they snuffed him out again as he tried the same incursion with Google F= iber, and realized it just cost too damn much to compete.=C2=A0 Cable Monop= olies, flawless victory.

Next I expect he'= ll team up with Elon or Bezos to try again via terrestrial.
<= br>
-mb


On Wed, May 6, 2020 at 10:32 AM M= ichael Butash <m= ichael@butash.net> wrote:
I tend to find the = CL network a bit wonky, having moved to DSL from Cox (damn bandwidth caps).= =C2=A0 I find the general performance is worse than cox, where I suspect th= ey simply don't manage the bandwidth and are far too oversubscribed as = it feels like the internet buffers at times, literally.=C2=A0 Cox would occ= asionally get that way too, and it was easy to see in an ongoing MTR when t= heir peering in LA would get slammed and latency would jump (not to mention= I know the guys that manage that bandwidth, telling them often got it fixe= d).=C2=A0 Oddly Using MTR with CL, they filter icmp/udp specifically that s= eems to hide responses to track well.=C2=A0 Go figure, truth hurts, so hide= it.

Having worked for service providers numerous = times over the years, working in and building them, routers are always an i= ssue in a metro city or even interstate networks.=C2=A0 No two platforms ar= e ever the same, whether buying all Cisco, Juniper, Nokia, or any combo of = all and more, which as you said, many do.=C2=A0 Hardest part is usually cap= acity planning, particularly with something like covid, every isp took a ki= ck in the groin at the same time to augment their networks, suddenly by som= e magnitude, when everyone else in the world is doing the same.=C2=A0 Slown= ess in networking can often be attributed to those not having enough capaci= ty, though they'll never admit it.

I'm= on the 150mbps dsl, and a speed test can provide that for sure, but genera= l usage, which I use a lot of tabs and apps, tends to bring things to a cra= wl often.=C2=A0 I'd even go back to cox if they got rid of the bandwidt= h cap.=C2=A0 CL might as well be government, and they're run by unions,= so nothing happens fast, including capacity augments.

Re: mac limits, having been around Cox both as a customer and netw= ork engineer working there early 2000's, the mac security was more abou= t limiting the amount of hosts behind a modem that could be allowed to a si= ngle mac and IP address.=C2=A0=C2=A0 Back Circa 1998 I had my first Cox mod= em, and there were no routers, you just got yourself a phat 10baset switch = from computer city and connected up your family on public ip addresses, eac= h with their own mac and ip's.=C2=A0 With no limits or filters that led= to security issues (hey, I see my neighbor's c drive shared!), Cox and= others then pushed people to then buy a router, which by then around 2002,= you could buy a cheap wrt54g linksys.=C2=A0 The advent of docsis also allo= wed to both filter and restrict the macs by default, also let them reduce t= o now 1:1 IP to User ratio, which was good for ip management, the abuse dep= artments, and fbi warrants from legal.=C2=A0 You used to be able to buy ano= ther ip, they'd push a new docsis config with mac-alowed=3D2, but not a= nymore.

Same reasons they're just building= in the router functions now, it ensures they can offer some basic customer= security, plus lets them run whatever spyware in their embedded router os = they want.=C2=A0 Better off buying your own standalone modem and router com= bo, one you ideally trust.

-mb


On Tue, May 5, 2020 at 10:07 PM Donald Mac McCarthy via PLUG-discuss <<= a href=3D"mailto:plug-discuss@lists.phxlinux.org" target=3D"_blank">plug-di= scuss@lists.phxlinux.org> wrote:
Putting a CL modem into a bridge mode where=20 it only handles the PPPoE connection is simply checking a radial select=20 button and hitting apply. If your firewall supports PPoE, even better,=20 as you no longer need their Modem and router in the mix. But, that is=20 just my experience, and it is limited. I have a CL fiber to the door=20 drop, and they gave me a Zyxel C3000Z device for connection. I promptly=20 ripped it out and allowed pfSense to maintain the PPPoE connection. I=20 had to call support for packet loss one time, and they refused to help=20 me. So goes it rolling your own I guess. Turns out a day later we had a=20 several hour outage due to one of the multiplexing cards used to=20 distribute the 40Gb/s core fiber to the GPON devices failed. Seems like=20 that was a likely culprit for some of the packet loss the previous day.

Having just gotten off a call in which the Senior Director of Security=20 Architecture and Engineering (a friend of mine from Atlanta) for Cox was a participant, before he hung up I asked him about the typical Cox=20 supplied modems. Very, very few of them are purely bridge devices -=20 especially with the push to "Panoramic WiFi". A member of CentryL= ink who was also on the call (ISP InfoSec sharing/working group) mentioned how=20 painful it was to support the number of company issued=20 modems/gateway/router models there are for different infrastructure and=20 connections - let alone ones that customers buy and bring to the party.=20 BTW, the MAC address thing is because they do actually use a MAC locking like feature for security. Apparently it is bad for the network if you=20 just go plug your modem in at several houses in the neighborhood due to=20 the way DOCSIS works. I still have to dig into that and ask some more=20 questions on that one.

There was a collective groan among the engineers when another ISP spoke=20 up about the number of critical flaws they find in their DOCIS devices=20 each year.

With the amount of consolidation which has happened in the past 20 years in the broadband market, the landscape is riddled with legacy bits and=20 pieces of this provider and that provider somehow being coerced into=20 working together to accomplish passing traffic. One of the ISPs=20 mentioned they had no less than 350 different models of core switching=20 equipment made by more than a dozen manufacturers in their network. They have a team of 40 (really 5 teams of 8) that simply monitor and ensure=20 that the OSPF functions properly among the various models and brands to=20 make sure that the network properly heals/manages congestion.

Anyway, just throwing it out so that people can see and understand the=20 picture at a higher level. The final comment on the call was from an=20 engineer at a midwestern rural provider and one that I am sure many of=20 us can relate to. She said she spends all day pulling her hair out=20 trying to keep the network functioning at the highest of levels. The=20 first words out of her kids' mouths when she gets home are "Mom, t= he=20 WiFi seems slow today."

I talked with Alexander this afternoon, and it looks like he has a=20 functioning network again. The APs were reluctant to give up their old=20 configuration, so a factory reset and new DHCP leases seem to have done=20 the trick.

Hopefully this sheds a bit of light on something for a few people.

Mac


Michael Butash via PLUG-discuss wrote on 5/4/20 4:59 PM:
=20
Ideally when you plug into a cable modem, it comes up, and passes your ethernet to the cmts in a bridge, lets one mac=20 address dhcp/arp, and things work.=C2=A0 It learns that one ip/mac, and=20 disallows any other mac.=C2=A0 No security, nat, nothing, just real dumb dh= cp + default routing with a public ip.=C2=A0 Routers/firewalls try to NAT you= ,=20 thus double NAT if using a router behind it.

C= L sells you a dsl modem/router that does your local security whether you=20 want it or not, full router/nat/firewall, and probably spyware.=C2=A0 Makin= g=20 it a modem is possible, but takes work, and your firewall has to support PPPoE (not all can/do).=C2=A0 Last time I touched a combo Cox router/modem= , I didn't see any way to do so.=C2=A0 I told them to buy a real modem, an= d that worked with their belkin/cisco/linksys/netgear they had.

If your "modem" mentions wifi, it's a router/firewall, not a mo= dem.=C2=A0 Not=20 all are clear about this, as they dumb it down for consumers, but an=20 important point.

-mb


On Mon, = May 4, 2020 at 1:53 PM Stephen Partington via PLUG-discuss <plug-discuss@lists.ph= xlinux.org> wrote:
I O= wned a Nighthawk Router/Modem combo, The way that=20 Netgear handled that is that the modem was hard-wired to a bridge on the router side. and technically you could see it as a separate=C2=A0device in= =20 the router configs if you rooted around enough. but the modem side was=20 just a modem.

On Mon, May 4, 2020 at 11:03 AM Michael Butash via=20 PLUG-discuss <plug-discuss@lists.phxlinux.org> wrote:
Cox modems *are* bridges first and foremost typically,=20 unless you get a bundled router/modem, which is only what CenturyLink=20 sells.=C2=A0 If you got a "router/modem" combo, just buy a modem-= only device=20 for a dumb bridge and simple ethernet for a public ip.=C2=A0 I recommend=20 staying with an arris cable modem, originally motorola, they basically=20 developed cable modem docsis, and are always the best.

I moved from Cox to CL when Cox started adding a usage cap, and that was=20 new to me to get my Fortinet firewall online with CL and their DSL doing PPPOE.=C2=A0 I've seen the router/cable modem combo boxes later, but n= ever=20 owned one as I always have my own router/firewall.

-mb


On Mon, May 4,=20 2020 at 8:36 AM Donald Mac McCarthy <mac@oscontext.com> wrote:
Wil= l Cox allow for a bridge/virtual bridge=20 mode? Xfinity does, which allows you to put in a firewall, and use the=20 modem only as a gateway, therefore preventing a double NAT situation.=20 Never lived in a Cox area before, and currently ride CL fiber.

Mac

Michael Butash via PLUG-discuss wrote on 5/3/20 2:00 PM:
Cox modems will learn and= =20 allow only 1 mac at a=20 time (unless business is set to allow more, but not on residential).=C2=A0 = If switching out firewalls, I 99% of time reboot the modem first and=20 foremost.

-mb

On Sun, = May 3, 2020 at 12:08 PM Snyder, Alexander J via PLUG-discuss <plug-discuss@lists.= phxlinux.org> wrote:
I got it working.=C2=A0

I assigned the SFP+ port as my LAN and assigned it the 10.x.x.x/16=20 network. Then I had to call COX and list the WAN Mac address with them.=20 Upon doing so I was able to reach external sites, and all downstream=20 devices started coming alive!

Thanks for all the suggestions and help!

Tha= nks,
Alexander

Sent from my Galaxy S10+

On Sun, May 3= ,=20 2020, 03:14 Herminio Hernandez, Jr. via PLUG-discuss <plug-discuss@lists.phxli= nux.org> wrote:
Can you=C2=A0 login=20 to the FW via the LAN interface? Can you=C2=A0 ping the FW LAN interface?= =20 Check the routing and NAT policy on the FW. All outbound traffic should=20 NAT to the FW WAN interface and there should be a default=C2=A0(0.0.0.0/0) route= to the internet.

On Sat, May 2,=20 2020 at 7:27 PM Seabass via PLUG-discuss <plug-discuss@list= s.phxlinux.org> wrote:
=20 I'm with Mac, I think it is not the firewall, but if you have the=20 ability to plug it into a display with a keyboard, you can use that for=20 configuration and modify a different device at the same time.

Makes it easier to troubleshoot by giving you the ability to configure your=20 pfSense ports at the same time.


Messag= e: 2
Date: Sat, 2 May 2020 09:04:35 -0700
From: Donald Mac McCarthy=20 <mac@oscontext.com>
To: "Snyder,=20 Alexander J via PLUG-discuss"
<plug-discuss@lists.= phxlinux.org>
Subject: Re: pfSense + Ubiquity
Message-ID: <= 18adfa38-3e72-7b0a-e31a-1ddf175d717f@oscontext.com>
Content-Type: text/plain; charset=3D"utf-8"

I can help - but I am unava= ilable to do so until tomorrow.

Make sure there are not any thing other=20 than default VLANs on the
interfaces to start with. Ubiquiti is=20 famous for not havinght eSFP+
ports active in the default=20 configuration, and I believe the switch has
all the ports to shutdown on default config as well.

I think it is the switch not passing=20 traffic through - no the firewall.

Mac
Snyder, Alexander J via PLUG-discuss wrote on 5/2/20 8:53 AM:
> Does anyone out there=20 have experience with pfSence and Ubiquity switches?
>
> I=20 have zero with either but that didn't stop me from buying both ....
= > how hard could it be?! LOL.
>
> I bought a Negate=20 XG-1537-1U. I bought a Unifi Pro 24 PoE switch.
>
> I can=20 configure the FW immediately after
>=20 firstboot/restore-default-configs, but only if i set the LAN interface
&= gt; to be the cable that goes directly to my laptop. That's great, but
= > that does shit for the downstream switch.
>
> I have a 10GB SFP+ Port that I want to configure as the downstream
> port to=20 ubiquity, but any configuration other than mentioned above
> fails .... and I'm now on my 12th "Reset To Factory Defaults" ... = any
> help on this would be greatly appreciated!
>
> Thanks,
> Alexander
>
> Sent from my Galaxy S10+
>
>
> ---------------------------------------------------
>=20 PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
> To subscribe, unsubscribe, or to change your mail settings:
>
https://lists.phxlinux.org/mailman/listinfo/plug-d= iscuss

--
Donald "Mac" McCarthy
Director, Field Operations
Open Source Cont= ext
+1.602.584.4445
mac@oscontext.com
https://oscontext.com
= -------------- next=20 part --------------
An HTML attachment was scrubbed...
URL: <http:/= /lists.phxlinux.org/pipermail/plug-discuss/attachments/20200502/aeab14b4/at= tachment-0001.html>

------------------------------

Sub= ject: Digest Footer

_______________________________________________
PL= UG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/plug-discuss


------------------------------

End of PLUG-discuss Digest, Vol 179, Issue 2
******************************= **************


--------= -------------------------------------------
PLUG-discuss mailing list -
PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailma= n/listinfo/plug-discuss
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailma= n/listinfo/plug-discuss
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss


---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/plug-discuss

--
Donald "Mac" McCarthy
Director, Field Operations
Open Source Context
+1.602.584.4445
mac@oscontext.com
https://oscontext.com
---------------------------------------------------
PLUG-discuss mailing list -
PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss


--
=
A mouse trap, placed=20 on top of your alarm clock, will prevent you from rolling over and going back to sleep after you hit the snooze button.

Stephen

---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss


---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/plug-discuss



--
Donald "Mac" McCarthy
Director, Field Operations
Open Source Context
+1.602.584.4445
mac@oscontext.com
https://oscontext.com
---------------------------------------------------
PLUG-discuss mailing list -
PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss


--
=
A mouse trap, placed on top of your alarm clock, will prev= ent you from rolling over and going back to sleep after you hit the snooze = button.

Stephen

---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss
---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss ---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss

--
=
A mouse trap, placed on top of your alarm clock, will prev= ent you from rolling over and going back to sleep after you hit the snooze = button.

Stephen

---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss ---------------------------------------------------
PLUG-discuss mailing list - PLUG-discuss@lists.phxlinux.org
To subscribe, unsubscribe, or to change your mail settings:
https://lists.phxlinux.org/mailman/listinfo/= plug-discuss --00000000000058ab1205a7cfe743-- --===============4271393506790542326== Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: base64 Content-Disposition: inline LS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tLS0tClBMVUct ZGlzY3VzcyBtYWlsaW5nIGxpc3QgLSBQTFVHLWRpc2N1c3NAbGlzdHMucGh4bGludXgub3JnClRv IHN1YnNjcmliZSwgdW5zdWJzY3JpYmUsIG9yIHRvIGNoYW5nZSB5b3VyIG1haWwgc2V0dGluZ3M6 Cmh0dHBzOi8vbGlzdHMucGh4bGludXgub3JnL21haWxtYW4vbGlzdGluZm8vcGx1Zy1kaXNjdXNz --===============4271393506790542326==--