- NANOG - lists.nanog.org

Google Cloud and IX - Traffic behavior
by Alain Hebert 21 Jun '17

21 Jun '17

Hi, Anyone aware of different traffic behavior depending if the target goes through normal peering than through an exchanges google exists in? We're facing a weird issue where the same GCLD Instance can upload up to 200Mbps (Ref 1) if the target path goes through, lets say TorIX, but cannot get more than 20Mbps on similar hosts (8 of them) sittings on our peering links. PS; Those sames hosts get up to their link limit ( 1Gbps ) between each others and others test points we have; PS: Wireshark capture show nothing abnormal; PS: Links aren't congested, and so on... Ref 1 - 200Mbps is on a link rate-limited to 300Mbps. Its my only test point with a TorIX access -- ----- Alain Hebert ahebert(a)pubnix.net PubNIX Inc. 50 boul. St-Charles P.O. Box 26770 Beaconsfield, Quebec H9W 6G7 Tel: 514-990-5911 http://www.pubnix.net Fax: 514-990-9443

4 7

Re: PCIe adapters supporting long distance 10GB fiber?
by Baldur Norddahl 21 Jun '17

21 Jun '17

But what foundation do you have for asserting that switch hardware is any different in this regard? I can say that we are using 80 km modules in various hardware without any issues. I admittedly do not use any high power modules in servers, but I will need better evidence than this to assume that it would not work just fine. Den 20. jun. 2017 22.24 skrev "Denys Fedoryshchenko" <denys(a)visp.net.lb>: On 2017-06-20 22:07, Baldur Norddahl wrote: > I would expect anything mounted in a computer to have all the power you > could want. It is not like the ATX power supply cares about an extra watt > or two. > > As I understand the issue it is more about cooling than power and is > primarly a concern in high density switches were you could have 48 or more > to power and cool. > SFP needs 3.3V, it might be supplied from regulator on the card or directly PCI-Express, can't be absolutely sure, in reference design it is just 3.3V_NIA and then filter, also reference design SFP power circuit define max 750mA/3.3V max to SFP, thats only 2.475W. FTLX1471D3BCV (10km SM) - up to 285mA FTLX1671D3BCL (40km SM) - up to 400mA, and thumb rule in electronics it is better to not exceed 50% of max specs of designed max current, as for many parts it is stated for 25C & etc operating conditions. I expect it might work, but noone knows how long, and how reliable, if it is not cooled very well. And 82599 sensitive to cooling(it is very old card after all), as soon as it is not enough, it starts to glitch. > > Den 20. jun. 2017 18.09 skrev "Denys Fedoryshchenko" <denys(a)visp.net.lb>: > > I guess it depends on NIC, there is many spinoffs of Intel X520 with much >> weaker power supply circuitry. >> It might work with good NIC, but you can't rely on it on long term, IMHO. >> Even 40km Finisar SFP+ has Pdiss 1.5W. Also they mention: "The typical >> power consumption of the FTLX1672D3BTL may exceed the limit of 1.5W >> specified for the Power Level II transceivers" >> If we talk about 80km, Pdiss is 1.8W. >> While 10GBASE-LR is <1W >> >> On 2017-06-20 16:30, Max Tulyev wrote: >> >> We use Intel NICs with SFP+ holes. It works good with long and short >>> range SFP+ modules, including CWDM/DWDM. >>> >>> On 15.06.17 12:10, chiel wrote: >>> >>> Hello, >>>> >>>> We are deploying more and more server based routers (based on BSD). We >>>> have now come to the point where we need to have 10GB uplinks one these >>>> devices and I prefer to plug in a long range 10GB fiber straight into >>>> the server without it going first into a router/switch from vendor x. It >>>> seems to me that all the 10GB PCIe cards only support either copper >>>> 10GBASE-T, short range 10GBASE-SR or the 10 Km 10GBASE-LR (but only very >>>> few). Are there any PCIe cards that support 10GBASE-ER and 10GBASE-ZR? I >>>> can't seem to find any. >>>> >>>> Chiel >>>> >>>> >>>>

3 2

Re: Vendors spamming NANOG attendees
by Scott Weeks 20 Jun '17

20 Jun '17

--- nanog(a)nanog.org wrote: From: i mawsog via NANOG <nanog(a)nanog.org> :: Agree, this thread has generated more "spam" or noise :: for all of us collectively. It's not spam. Look up the definition of spam. Also, just block the thread in your email client. :: Some amount of relevant "spam" has to be tolerated for :: vendor to continue supporting NANOG. No, that is absolutely wrong! :: Also relevant "spam" or sales call is a good way to find :: out about new technologies , that one may not have heard :: about otherwise. No, wrong again. One of the most attractive things about NANOG is learning these things from **peers**, not spam. :: Another tip, just ignore, the "spammer" will go away :: eventually. If they're not named-n-shamed they will continue. I have had quite a few go after me aggressively to sell me something after a NANOG post. They didn't stop until I named-n-shamed them. Their supervisors wrote me, personally apologized and, due to that, I put their companies back on my list of companies to do business with. I write this not just to you, but as an explanation to other new folks why the old schoolers name-n-shame. scott

1 0

Re: Vendors spamming NANOG attendees
by tim＠pelican.org 20 Jun '17

20 Jun '17

On Tuesday, 20 June, 2017 14:26, "Rod Beck" <rod.beck(a)unitedcablecompany.com> said: > And how do you tell if an address was scraped or not? There are databases and > zillions of other ways of gaining addresses. > > > I doubt you can distinguish the source with any real reliability. Depending on whether you're registered with personal or corporate email, and how much control you have over the platform in question, you can distinguish the source with fairly high reliability. Just generate a new 'bob+nanog70(a)bobsdomain.org' style address for every event you register for, every website that requires a contact address, every mailing list, ... If you're concerned that people will twig, and use the naked 'bob@' address, you could work with multiple names including a hash that look like internal nonsense, e.g. 'bob34adf@', or block the un-plussed 'bob@' entirely and use e.g. 'robert@' for people you trust to have your real, non-circumstance-specific email address. I know people who do this, it really depends how much you care about being able to trace and block people who are either scraping or re-selling your details. Regards, Tim.

4 4

Vendors spamming NANOG attendees
by Rodney Joffe 20 Jun '17

20 Jun '17

It seems that more than just a few of us were spammed by Glenn Stern (gstern(a)calient.net) an employee of Calient following NANOG 70. The spammer had the balls to say, in his email: > > We do not know each other. I'm leveraging the attendee list for NANOG to reach out and raise awareness of the value of OCS (Optical Circuit Switching) in the data center and in particular, the Carrier Neutral Hotel where we've been active with next generation MeetMeRoom discussions. He does not show as an attendee at NANOG, but another executive, David Altstaetter, daltstaetter(a)calient.net did register, and may have even shown up. Hopefully those of you who have traditional community attitudes will show your reaction via your pocketbooks. Maybe its time for the NANOG board and staff to step in, and develop some teeth to use in cases like these? Unless the majority of you members are cool with unfettered spamming of member and attendee lists. In which case, have at it! Rodney

38 67

SAFNOG-3: Sponsorship Opportunities
by Mark Tinka 20 Jun '17

20 Jun '17

Hello all. It gives me great pleasure to announce that the Sponsorship Matrix for SAFNOG-3 is now available at the link below: http://www.safnog.org/matrix.html If you are interested in collaborating with SAFNOG-3 as a sponsor, please do not hesitate to reach out to: secretariat at safnog dot org Any and all support will be sincerely appreciated. We look forward to seeing you in Durban for our next meeting. Cheers, Mark Tinka On Behalf of the SAFNOG Management Committee

1 0

[NANOG-announce] Mailman Maintenance - June 19 - 9PM Eastern
by Valerie Wittkop 19 Jun '17

19 Jun '17

Hello NANOGers, The mailman servers will be undergoing a second maintenance tonight between 9:00 and 10:00pm Eastern. Mail sent to lists hosted on the NANOG mailman servers will be held during the maintenance window. Regular operations will resume once the maintenance is complete. Sincerely, NANOG Staff _______________________________________________ NANOG-announce mailing list NANOG-announce(a)mailman.nanog.org http://mailman.nanog.org/mailman/listinfo/nanog-announce

1 0

Re: DWDM Mux/Demux using 40G Optics
by Tony Wicks 19 Jun '17

19 Jun '17

The guys at fibrestore will point you in the right direction on all this if you ask them these questions. They are actually very helpful and will assign you a specialist to assist. -------- Original message -------- From: Colton Conor <colton.conor(a)gmail.com> Date: 20/06/17 6:26 AM (GMT+12:00) To: NANOG <nanog(a)nanog.org> Subject: DWDM Mux/Demux using 40G Optics We are building a 40G metro ring using 40-Gigabit Ethernet QSFP+ Transceivers. Specifically, we are using Juniper JNP-QSFP-40G-LR4. This is a QSFP+ Transceiver with a LC duplex head. We only have one pair of single mode dark fibers around the ring. Our distance between nodes around the ring are all less than 10KM, so we can use standard optics. We go out of one JNP-QSFP-40G-LR4 and into another JNP-QSFP-40G-LR4. There are no passive muxes involved. This is working great for 40G. My understanding is a JNP-QSFP-40G-LR4 is really a transceiver with a CWDM mux built into it. The spec sheet shows it sends 4 10G channels: https://www.juniper.net/documentation/en_US/release-independent/junos/topic… Lane wavelength Lane 0–1264.5 nm through 1277.5 nm Lane 1–1284.5 nm through 1297.5 nm Lane 2–1304.5 nm through 1317.5 nm Lane 3–1324.5 nm through 1337.5 nm This setup is working fine, but now we want to do more than 40G around the ring. To my knowledge there are no other 40G QSFP+ transceivers that use four other channel/lanes than the ones already being used, so they only way to go higher than 40G is to stack 10G or 100G channels ontop of the fiber pair using a passive mux. 100G is too expensive for the time being, so we are looking to add 10G channels to a ring that already have one 40G channel using the QSFP+. I was reading this tutorial, and it mentions "there is a 1310 nm port integrated in a 40 channels DWDM Mux/Demux system. The 1310nm added port is a Wide Band Optic port (WBO) added to other specific DWDM wavelengths in a module. When we run out of all channels in a DWDM Mux/Demux system, we can add the extra optics via this 1310nm port." http://www.fs.com/upgrade-to-500g-with-40ch-dwdm-mux-demux-system-aid-493.h… What I can't seem to understand is they are mentioning that this 1310 port can pass QSFP+ signals, so it sounds like its really a 1270nm through 1330nm port? Is this what they mean by Wide Band Optic port (WBO)? We don't need 40 10G channels plus a 40G for a total of 440G. More than likely we are looking at a 8 channel mux/demux, and 1 40G port for a total of 120G. I don't care if we do CWDM vs DWDM, but I assume it will be hard to find a CWDM mux that has one LC dupluex input for 1270nm through 1330nm channels? Maybe I should just ditch the 40G QSFP+ optics and use all 10G optics, but the switches I am using have 48 10G SFP+ ports and 6 QSFP+ ports built in. I know there are 40G breakout cables, but the whole point of 40G is to aggregate VLAN/circuits. Has anyone done this before?

2 1

test - ignore
by i mawsog 19 Jun '17

19 Jun '17

test - ignore

1 0

GozNym - Gozi ISFB - Nymaim
by Aaron Gould 19 Jun '17

19 Jun '17

Anyone experienced in stopping/blocking/cleaning GozNym - Gozi ISFB - Nymaim ? - Aaron Gould

2 2