Video

Hat dieser Hackathon eigentlich mal stattgefunden für 25Gbit CPE's? Oder ist die "Lösung" einfach das was Stapelberg in seinem Blog geschrieben hat?

Segment Routing - The best way of explaining new feature of MPLS technology.

Nice presentation 👍

Finally, IPv6-only pfsync ^^

Great practical introduction to containerlab, Jens!

Great presentation.

Schee! 😊

It has been a great event. Looking forward for #DENOG15

This is the best primer on module creation ever!

Aviate - Navigate - Communicate - Another important rule in aviation. Fix Your network first, Think about options, Communicate The problem

Great to hear from ISC.

But i must make a good plan for the community-numbers ans that meanings😅

Greatest video over BGP Communities that I watched, i must probe it tomorrow

can you please help to provide guidance for IXP setup?

Oliver Knapp is wrong at several places. The C-VLAN is not a "Customer-VLAN" but instead a service vlan (VOIP, IP, MULTICAST, TV, WHATEVER). N:1 and 1:1 are totally different aggregations. Most of what he said in the middle of the talk is wrong.

Excellent presentation!

It's a bit too abstract, but he stated it was not a tutorial. Good job.

Nice talk!

Great talk Pim. Too bad you talk so slow ;)

awesome - well done guys

"Vincentz - Hat of Network" :)

Spannender Einblick. Wusste nicht, dass das mit den "Titeln" in corporate so ernst genommen wird.

Thank you

Great video, thanks for doing this topic!

very nice

Init7 ❤️

Great Presentation. Thank you!

Traurige Deutschengeräusche.

the Image is very bad!

Ich erinnere mich noch an die Berichterstattung. Dass das so knapp aus dem Boden gestampft wurde, ist der eigentliche Skandal. Gut gemacht, super erklärt!

Hi. I'm trying to use Blaster to simulate PPPoE sessions, but it's failed. BNG Blaster sends PADI, and BNG already replied PADO, but after that, BNG Blaster did not send more packets to establish session. Could you help me to check why ?

Awesome!!!! Thank you!

👍

One aspect of blameless post mortems is to go out to those who make mistakes, get introspection, and recommendations as to what can prevent others from making the same mistakes. In other words, the attitude I always took in post mortems was that if you make a mistake that is human, but if you can make the mistake, others can and therefore who was closest to the mistake has the most valuable perspective in how to keep others from making that mistake. One challenge is getting junior folks to adopt this idea that mistakes are opportunities to do better not just individually but collectively. It would be interesting to get your view on what can be done to cultivate this.

I noticed you didn't cover structured decision models. One of the really cool things about aviation decision making is that the decision models are collaborative and foster improved communications. This is another area we really can learn in IT from aviation.

There are some cases where some pieces of airplanes are still not redundant. There are non-redundant parts in trim control systems for example. But these have extra certification requirements and inspection requirements. So there are ways of compensating for non-redundance in IT too.

An important point in the two pilot rule is also division of roles between pilot flying and pilot monitoring. This division of tasks is super-important too. The second operator in your view can also be checking monitoring periodically, scan for unexpected behaviors etc.

An important part of checklists though is workload reduction and this is a *big* reason we need checklists in IT. For example, if you are responding to an alert at 3am, having a checklist that does not require scrolling makes it far easier to follow steps and not forget anything. And I mean "workload" here in the aviation sense in terms of number of things you are facing at the same time. On the sterile flight deck rule you have the same point you are making, which is to divide tasks so that people have to pay attention to fewer things at the same time.

das beste beispiel fuer unglaublich viele dieser faktoren ist immernoch die "grosse gitlab outage": man scheitert oft beim beherrschen der systeme und kommt so nicht an den luftfahrt standard - aber all die surrounding sachen hat man zur verfuegung und kann sie im Griff haben, so eben z.B. uebermuedete Leute nicht alleine lassen, standardisierte Verfahren fuer Checks wie "bin ich auf dem richtigen System?", standadisierte Erkennungszeichen fuer kritische Systeme etc. Mein aha-moment war vor vielen jahren, als ich anfing bei change requests die "rollback" punkte wirklich auszufuellen und -denken. leider wurde das nicht gross aufgegriffen damals. fuer failure exercises: beim veritas cluster gab's nen fire drill feature - wir haben in 4 jahren, wo ich damit gearbeitet hab, nicht EINMAL durchbekommen, dass wir das probieren durften. bei nem config management system hab ich es direkt geplant, sowas zu machen und es scheiterte daran, dass ich halt unbedingt effektive gruppenzusammensetzungen fuer die fire drills empfohlen hab. fuer meetings? da geht das IMMER. aber doch nicht fuer sicheren betrieb! was Du ueber Decision making gesagt hast, ist sehr interessant und war mir neu - der Vergleich zur Schifffahrt, auf jedem Schiff ein Alleinherrscher, und die vielen Ungluecke durch abwechselnde Korrekturversuche draengt sich auf.

We are an MSWe are an MSP and we starting a new project here in Brazil - Building 4 new edge datacenters. Our focus will be VPS and Collocation and we dont have availability of IPV4 blocks in Brazil anymore. You vídeo help us to understand more the possibility to use only IPV6,. Thanks

Great talk Mel!

Ich frag mich mittlerweile halt, wofür man diese ganze Bandbreite braucht. Drum können die "grossen" Provider ja tausendfach Oversubscription betreiben.. jeder will Gigabit haben, und dann hängt er den ganzen Tag auf Facebook und CZcams rum. Ich habs bei meinem letzten Upgrade von 200/50 auf 500/100 (TV-Kabel) schon gemerkt dass das eigentlich rausgeworfenes Geld war.. ob meine Steam-Updates jetzt 10 oder 30 Sekunden brauchen, macht eigentlich keinen praktischen Unterschied mehr. Gerade im Hinblick auf Homeoffice und Cloud Storage wären symmetrische Verbindungen wichtiger als immer mehr Downstream.. aber das gibts halt leider nur mit Glas, und Glas gibts noch viel zu wenig.

Ah, super spannend. Gut gemacht und danke für die Einblicke.

I really like pfSense, but after I learned how vicious and vile Netgate employees and pfSense fanboys reacted to the OPNsense fork, I'd rather not have anything to do with them.

Just beat u on Warzone LOL GGs

I would love to switch over to IPv6 completely. Sadly I got an O2 LTE contract and only realized afterwards that O2 will introduce IPv6 right after hell froze. So I have to stick to dual stack for at least the next 1.5 years.

Thank you guys... Good job

fantastic talk

Sehr informativ und echt cool mal eine Zusammenfassung von dem ganzen Thema zu bekommen! Weiter so init7 👍

Thanks for recoring and uploading !