Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
Networking Education The Internet IT

Take a Free Networking Class From Stanford 128

New submitter philip.levis writes "Nick McKeown and I are offering a free, online class on computer networking. We're professors of computer science and electrical engineering at Stanford and are also co-teaching Stanford's networking course this quarter. The free, online class will run about six weeks and is intended to be accessible to people who don't program: the prerequisites are an understanding of probability, bits and bytes, and how computers lay out memory. Given how important the Internet is, we think a more accessible course on the principles and practice of computer networks could be a very valuable educational resource. I'm sure many Slashdot readers will already know much of what we'll cover, but for those who don't, here's an opportunity to learn!"
This discussion has been archived. No new comments can be posted.

Take a Free Networking Class From Stanford

Comments Filter:
  • IPv6 (Score:5, Interesting)

    by AK Marc ( 707885 ) on Sunday October 07, 2012 @02:30PM (#41577767)
    Just a quick glance at the syllabus, and I didn't see anything about IPv6, if IPv6 gets deployed in the next 4 years, wouldn't that make the intro course essentially obsolete before anyone finished a degree? Remember, this is the year of IPv6 on the desktop.
    • Re:IPv6 (Score:5, Insightful)

      by rogueippacket ( 1977626 ) on Sunday October 07, 2012 @02:38PM (#41577819)
      I would agree with you if it were a 12 or 16 week course - but there is a lot to talk about in terms of the physical and data link layers and plain IPv4 before even addressing IPv6. Six weeks is barely enough time to provide a solid foundation there.
      • Except the lower layers typically aren't your problem, or require much "design". You buy 10G equipment from a reputable vendor, and you dont worry about it. What IS important is layer 3 and layer 4 stuff - that's where people fail all the time. "can you make a DNS entry for a webserver on a different port?" "its easier if you just use the 169.254 subnet" etc.

        • by AK Marc ( 707885 )
          Especially if one thinks of the "complex" parts of L2 in L3 equivalence. STP sometimes seems to be RIP, based on MAC rather than IP. So if you know RIP well, STP is an easy learn for the concepts, though not some of the tweaking and priorities. Most of the L2 issues are similar to L3. And L2 is often an extension of L3. IGMP snooping and things like that make it silly to learn up the OSI in a linear manner. Why would somone learn IGMP snooping before learning IGMP?
          • by AlphaWolf_HK ( 692722 ) on Sunday October 07, 2012 @07:20PM (#41579715)

            I disagree. If you don't learn much of the layer 2 concepts, then you'll probably never learn anything about layer 2 troubleshooting. One time I had to work on a network where they had separate discontiguous VLANs that couldn't talk to one another. I remember a while back hearing about a major hospital whose network failed because their engineers didn't understand some of the more advanced layer 2 concepts. There are also issues such as implementing measures preventing some idiot from creating switching loops that STP can't detect, e.g. cascading some $10 switches they found at wal-mart which are creating a layer 2 broadcast storm that is bringing down an entire VLAN.

            Troubleshooting isn't the only thing, you also won't understand layer 2 security, which has been exploited quite a bit lately. Some easy examples would be MAC flooding and MAC spoofing, to the more severe problems like VLAN hopping, and much more that I can't think of off of the top of my head.

            • by AK Marc ( 707885 )
              VLAN hopping is not unlike IP spoofing. You are disagreeing, while proving my point. The "complex" tasks in L2 generally have an analogue in L3. So L2 is of lower importance. And nobody uses L2 security, other than wireless. 802.1x was "invented" for wired networks, but almost no one used it for anything until WPA. And generally, wireless is treated as a separate subject.
              • No, not really, in fact I would say not even close.

                IP spoofing doesn't give access to portions of the network that are blocked by layer 3 filters, and I mean properly done IP filters that will block address ranges that couldn't have come from the source interface (be it a physical one or SVI.) Even in the rare circumstances that this can be done, often times the routing tables won't handle the return traffic. Not only that, but IP spoofing can't give you access to networks that are logically isolated entire

                • by AK Marc ( 707885 )

                  IP spoofing doesn't give access to portions of the network that are blocked by layer 3 filters, and I mean properly done IP filters that will block address ranges that couldn't have come from the source interface

                  And how does that differ from a properly configured port not allowing VLAN hopping?

                  Unlike IP spoofing, VLAN hopping can do all of the above, unless of course you wanted to go back to pre 2001 technology and do away with layer 3 switches, and even do away with VLAN's entirely.

                  You are making it sound harder than it is.

                  No matter how you slice it, you should NOT ignore layer 2.

                  Fuck you. I'm done. You are wrong. And you are lying. I never said someone should ignore layer 2. If you are going to lie to make your points, I can't think of any response short of "fuck you".

                  I'm not trolling or trying to be rude, that's just reality.

                  The reality is that, unless you quote a post where I said "You should ignore layer 2" then you are a lying troll who is lying to be rude to prove some point that nobody but you cares about.

                  • And how does that differ from a properly configured port not allowing VLAN hopping?

                    You are making it sound harder than it is.

                    I'm approaching this from the angle that you don't learn much about layer 2, as you suggested earlier. If you don't learn much about it, how are you supposed to know to watch for this (among other layer 2 security issues) to begin with? And if you don't know to watch for this, then how are you supposed to learn go configure it as well?

                    Fuck you. I'm done. You are wrong. And you are lying. I never said someone should ignore layer 2. If you are going to lie to make your points, I can't think of any response short of "fuck you".

                    Well the parent post to yours said "don't worry about layer 2," and you seemed to agree with it. Sorry if you don't like that interpretation, but "don't worry about it" and "i

      • Re:IPv6 (Score:4, Interesting)

        by kasperd ( 592156 ) on Sunday October 07, 2012 @04:27PM (#41578531) Homepage Journal

        there is a lot to talk about in terms of the physical and data link layers and plain IPv4 before even addressing IPv6.

        It would be better to teach plain IPv6 before you start addressing IPv4.

      • by AK Marc ( 707885 )
        They mention routing. Understanding routing requires understanding the history of routing. CIDR is a meaningless term if you just learn CIDR from day one, as you would have to learn classes to understand the distinction of classless, so covering CIDR would be a redundant history lesson, followed by why we ignore the history now, and that classes and CIRD are irrelevant with IPv6.

        So I'm curious how, 10 years from now when we are all IPv6 native and IPv4 is used only by child porn pushers, how classes/CIDR
        • by kasperd ( 592156 )
          You can explain how the hierarchy of networks and prefix lengths work without having to explain the entire history of classes and CIDR. With IPv6 I don't think you'll hear people talking about class A, B, and C networks, because the prefix lengths for those are only defined for IPv4. If you are teaching this stuff, you'll have to decide if you are teaching a history lesson or if you are teaching how it works. In the later case a lot of history should be left out.
          • by AK Marc ( 707885 )
            In a practical sense, if you were interviewing a "network expert" who didn't know what "class A" meant, would you trust them? It seems like a core piece of knowledge. When you teach RIP v2, do you just pretend RIP v1 didn't exist? How do you explain why RIP v1 was mostly unusable in most networks due to classes being required?

            You can teach RIP v2 and let everyone assume there was some older RIP that's unused, but curious students should ask why. It's hard to teach anything without teaching history. Th
            • All network experts don't have to be experts in IPv4, since that is starting to dry up. If you are trying to build an IPv6 network from ground up, it would actually help that they don't have the ball&chain IPv4 thinking constraining their ideas of what is possible.

              Having said that, while there are no classes as such in IPv6, you do have some ISPs offering anything from /48 to /64, which is why in Windows 7, in the IPv6 properties, one is prompted for the number of subnets. Each subnet would still ha

            • by kasperd ( 592156 )

              if you were interviewing a "network expert" who didn't know what "class A" meant, would you trust them?

              As a matter of fact, I have interviewed plenty of people on networking. To this date, I don't know if any of them know what class A meant. I asked them questions, which I considered more important. If I had come across a candidate, who knew nothing about IPv4, but did have the level of expertise I was looking for in IPv6 instead, I would have recommend he be hired. But in reality I might never have found out, as I was mainly asking questions at a higher level.

      • Why is teaching IPv4 necessary at all? Particularly for a free course? Teach IPv6 independent of it, and in a separate paid course, teach things like transitional technologies b/w IPv4 and IPv6.
    • Re: (Score:3, Interesting)

      Not even Cisco teaches IPv6 in their shitty little Network Academy (they also teach very little OSPF, and are really big on their proprietary EIGRP).

      I can't think of any basic networking class that teaches IPv6.
      • My CCNA class skipped IPv6 in favour of extra security stuff because 'no-one cares about IPV6'.

        I've since done some HP Networking training and exams. They're better on teaching open standards; LACP rather than etherchannel, LLDP rather than CDP, etc. However, the material still glossed over IPv6. There were very few questions on it in the exams, and if it's not in the exam, people won't learn it.

      • I've learned almost all I know about networking from the Cisco Network Academy offered at a local community college. Part of the course material is to learn OSPF pretty extensively, even going as far as to explain the differences in how it handles single access or multiple access, and in doing so how the DR and DRO elections work.

        There is a fair amount of coverage on EIGRP, but I would say they teach both about the same. The teacher plainly told me that you never see EIGRP in a mixed environment. However on

      • I thought that the current CCNA is focussed more on IPv6 than IPv4. Isn't that the case? If not, I'm glad I didn't take the exam.
      • Well, there is an IPv6 Education Certification Program [infoweapons.com], that should cover everything from an IPv6 POV.
    • shows how 4 year degree are not for IT / desktop / sever / network work.

      and why IT needs to be tech / trades like with out being tied to a college time table. And having teachers who are not in school for life.

    • by fm6 ( 162816 )

      IPv6 is already deployed. Its on my Windows laptop (and has been for a couple years) and is supported by my ISP. But knowledge of IPv6 doesn't become essential until substatnial bits of the Intertubes stop using IPv4. That might well be more than 4 years away.

      And even when that happens, it might well make sense for an introductory course to concentrate on a more simple model that beginners can more easily understand. I have a friend who teaches introductory assembly language as a way of helping CC students

      • Re:IPv6 (Score:4, Interesting)

        by kasperd ( 592156 ) on Sunday October 07, 2012 @04:23PM (#41578507) Homepage Journal

        it might well make sense for an introductory course to concentrate on a more simple model that beginners can more easily understand.

        In that case teach IPv6 and skip the parts that nobody use. IPv6 is a little bit simpler than IPv4. There is not a huge difference, but there is certainly no point in teaching an obsolete technology for simplicity, when it isn't simpler. IPv4 is not entirely obsolete yet, but judging from the number of people who think IPv6 is more complicated than IPv4, I perceive that there must be a shortage of people who understand IPv6.

    • No, it wouldn't. IPv6 deployment does not mean IPv4 will disappear. There are a lot of market and backwards compatibility reasons (e.g., the monetary value of IPv4 addresses). The course's coverage of IPv6 is mostly in the realm of names and addresses. We don't go into the mechanisms (such as ND, RA, etc.), mostly because it is much easier for students to interact and observe IPv4 networks than IPv6 ones. Generally speaking, for introductory material like this it's better to teach about a certain today than
    • I fully agree w/ this. Given that a lot of paradigms in IPv6 are new and have nothing to do w/ IPv4 e.g. subnetting, various scopes of addressing, particularly in multiplexed addressing, link local vs site local addresses and so on, I think that newer networking courses should offer pure IPv6 courses, and then separately delve into any interoperability issues w/ IPv4. Right now, a lot of engineers' are tunnel-visioned into IPv4 based limitations, and shouldn't have to deal w/ it. Particularly, in a netwo

  • by betterunixthanunix ( 980855 ) on Sunday October 07, 2012 @02:31PM (#41577771)
    I love the concept, but there is a problem: your prerequisites will exclude the overwhelming majority of non-programmers, and might even exclude a few novice programmers. Probability and how memory is laid out in a computer are the biggest issues here; even many computer science programs do not cover these topics until the second or third year.
    • by DamonHD ( 794830 ) <d@hd.org> on Sunday October 07, 2012 @02:42PM (#41577857) Homepage

      Probability should be taught early on at school, not waiting for university!

      Rgds

      Damon

      • You would think so, but let's put it this way: a large number of the undergrads I have met are baffled by the Monty Hall problem, and that includes those who have actually taken probability in college. Probability is just not taught very well these days.
        • Re: (Score:2, Insightful)

          by Anonymous Coward

          That's a pretty bad example. The Monty Hall problem is often disputed even among the very well educated, and for a time that even included Paul Erdos (https://en.wikipedia.org/wiki/Monty_Hall_problem).

          Being baffled by the Monty Hall problem would not seem to be a reliable indicator of poor probability education. It just shows that the Monty Hall problem, for most people, isn't intuitive *despite* being educated.

          • The Monty Hall problem is often disputed even among the very well educated

            No it isn't.

            It just shows that the Monty Hall problem, for most people, isn't intuitive *despite* being educated.

            Most things to do with probability aren't intuitive, that's the whole point in studying it. If human brains worked so that they cold analyse probablity as easily as catching a ball or walking, we wouldn't have to teach it at all.

            I will never forget when I first came across the "how many people do you need in a room for it to be virtually certain two will share a birthday?" problem. I now "know" the answer, but it certainly doesn't seem intuitively right to me even now.

      • Well when I went to school probability didn't even exist. :)

    • My thoughts exactly...Also, why would knowledge of these areas be a prerequisite for a networking class to begin with?
      I'm not being snide, I'd actually like an answer if possible.
      • Well probability will be needed for information theory, error correcting codes, and cryptography. I am not sure if that is how detailed this course is going to get though.

        As for the layout of memory, I would just guess that they might talk about security and e.g. buffer overflows.
        • That sounds like a bad idea, that strays a lot from networking. Not to mention the fact that if someone has memory management/layout related knowledge he'll already know about buffer-overflows, ROP and other related security matters and will most likely also have a decent amount of programming experience (unless they meant a "looked at a memory layout diagram once" level knowledge).
          I'm curious now...I may sign-up just to see how they'll weave these different topics together in their material.
        • All of that goes beyond the realm of networking to be honest, with the slight exception of cryptography. As far as cryptography goes, you don't bother with that until you are at a more advanced level, and even then, you only need to be familiar with the names of the algorithms and the general purpose of each.

          Most of what you are talking about goes in the area of computer science, and networking doesn't necessarily include computer science, and likewise, computer science in itself doesn't necessarily include

      • that is the college line of thinking loads of theory that for most people is not needed or is off base too top level.

        Also all that knowledge is too much one size fit's all where networking can be it's own class track with people needing different levels of it based on what they are working with.

    • Re: (Score:3, Informative)

      Talking about packet formats is difficult if the student doesn't understand the idea of bits, bytes, and memory addresses. How memory is laid out isn't mean to imply segments, pages, virtual memory, caches, registers, or anything like that, just the idea that you have cells of bytes.

      Looking forward, I agree with you -- it would not be hard to figure out how to present some background material on these topics to make the course more accessible. But this is just the first time we're teaching it. We'll lear
      • by Phylarr ( 981216 )

        My (in my opinion perfectly capable) wife wanted to take this course but decided against it after reading those prerequisites.

        It will take more than just free online courses for academia to overcome its walled garden and truly be more accessible. Why is something given as a prerequisite that can just as easily be explained in a couple of minutes in the first lecture?

        • Why is something given as a prerequisite that can just as easily be explained in a couple of minutes in the first lecture?

          Self-selection by potential students will mean that only people who know a lot about the subject or related subjects already will apply. They will then find it easy to follow, and give the college lots of great free publicity when they comment about how the course made them an expert in a couple of hours and was full of great teaching, even though they knew half the stuff already..

    • I haven't read any details about the Stanford one yet, but the Cisco Network Academy I went to starts you from scratch. No need for programming, no need for understanding memory (though I am a novice programmer, the other students who had no knowledge whatsoever picked it up fine.) By the time you're done, you can easily count in binary up to 8 bits in your head.

    • I can see non-programmers like me fulfilling the first few of them, but something tells me the probability he's referring to isn't what we used back in high school or college math, and I have no idea where we would have learned the layout of memory outside of a programming class. When I took a college C programming class (which I barely passed & didn't learn from), the instructor said that handling RAM was taught in the "advanced" C class, as we needed to master the basics first; I similarly didn't see

  • Non-programmers... on Slashdot? Did the demographics of the user base here change when I wasn't looking?
    • Non-programmers... on Slashdot?

      What do you think IT people are?

    • by xclr8r ( 658786 )
      Not all engineers are programmers, not all Sciences involve programming. Slashdot has had a gaming section for a long time and definitely not all gamers are programmers. The demographics have changed but not as much as one might think. I'm actually thinking that those science fields are picking up more on programming than use to be the case.
    • Keep in mind that "news for nerds" does not just mean "news about things that nerds are paid to do;" it also means "news about things that affect nerds or might otherwise be of interest." Considering how poor the general understanding of computers and computer networks is among the majority of the population, and the fact that the poor understanding of technology leads people to institute or accept policies and to make choices that Slashdot readers typically think are bad, a course that is supposed to help
    • There are probably more sysadmins than pure developers on Slashdot. I suspect this was always the case.
    • Non-programmers... on Slashdot? Did the demographics of the user base here change when I wasn't looking?

      Despite what some people here like to think, there are other jobs and interests in life than being a full time software developer that still mean you fall under the nerd umbrella.

  • by tomhath ( 637240 ) on Sunday October 07, 2012 @02:49PM (#41577917)
    I suspect most would benefit from an introductory course.
  • As usually said on woot "In for one".

    Should prove interesting. I've been a network engineer since 94 and I'm quite honestly disgusted by people who claim to be network engineers these days who don't understand the difference between Bit/Byte, the concept of a Packet or what CSMA/CD is. Administrators I'd expect that from, they don't really need to know these things. Engineers should, IMO. These are the basic building blocks of "traditional ethernet" networks, how can I expect someone to actually design some

    • I don't know about anybody who claims to be a network engineer these days other than my peers, and all of us have been through CCNA, and you can't pass CCNA if you don't know that, and a lot more.

      • by Raxxon ( 6291 )

        Last time I looked at the CCNA courses I considered it "baseline" stuff (this being back in 99/00 timeframe). But I know a number of people who say the words "Network Engineer" and yet the only "engineering" they've done is to hire a cabling contractor to lay cables and then plug them into a few switches. The phrases "Network Map" and "Site Bible" mean nothing to them. There's no documentation of the design/implementation of the network, there's usually no disaster plan, half the time a simple stupidity of

        • Part of the class grade was that you had to document very well everything you did in the class labs. Not just physical and logical topology diagrams, but also a writeup of what configuration choices you made and why.

          I've heard from some people who I've shown what I've done that say that what I learned was formerly only CCNP and even some CCIE material, meanwhile CCNP and CCIE have advanced much further.

          Supposedly back in 2007 there was a major revision to the CCNA course material, and before that another on

          • by Raxxon ( 6291 )

            Looks like I need to reconsider the certification then. Last I honestly looked at it it was on par with Novell's CNA and Microsoft's early MCSE.. Great for getting an introduction, but not something that's going to really take you far.

            I'm still finding myself wishing at times that the CNX (Certified Network Expert) program hadn't fallen apart, but when you're a program governed by various (and opposing) members of the industry I'm shocked it got out of the womb....

  • You didn't have to do this. With Stanford's name behind you, you might even have made a fair chunk of cash selling it. But you didn't.

    Thank you.

  • Will this course teach enough to be useful for those who want to go into networking on a professional basis? For instance, will it enable them to pass the first of the Cisco network certifications? Will it confer enough skill and knowledge to do anything practical? Give credit or fill a prerequisite in a program that will go that far? Or is it just a feel-good class?

    In my experience with Universities, "FOO for non FOO majors" and other survey classes have impressed me as snow jobs, shoveling out a lot of material in a disconnected and difficult to absorb way. They seemed directed more at giving non FOO majors an increased level of respect for specialists who have mastered the subject than to confer any useful skill or knowledge. I'm hoping this isn't another of the form.

    • A Cisco certification is useful for someone who needs to pick up practical skills immediately and become useful in a job right away. This Stanford course, however, is probably more useful for someone who's been in the industry for sometime and wants to understand networking on a deeper level.
    • I don't think that's what this would be aimed at. Most of your time as a network engineer will be focused mainly on layers 2 and 3, with a bit of 4 and even less as you go up to 7. From reading the description page, it looks like they are going into layers 6 and 7 rather extensively, (namely talking about HTTP and bittorrent) though I couldn't say for certain without actually going through the course.

      • That is typical of what you get when computer science people teach networking classes. It is certainly not the case that all computer scientists and programmers look down on infrastructure people and concepts...but a lot of them do...you'll find that the case on Slashdot as well. People who program don't see the level of training in theory and practice that is important for network engineers and (good) system administrators. They just don't see low level protocols as interesting. Of course, my experience is
    • by HnT ( 306652 )

      Very likely NO on the Cisco certification. I was a lot into networking during university and ultimately did a lowly CCNA on the side and I can guarantee you that no training but a Cisco-specific, if not CCNA-specific training will get you through even those entry level certifications.

      While the networking theory alone should be enough, the cisco, router and IOS specific questions will break your back if you haven't prepared well for them. e.g. expect to be asked situational, specific IOS commands in multiple

  • Stop creating kewl online courses that I just have to take faster than I have time to actually take them!

  • I applaud your efforts to teach this class, and I just enrolled.
  • ...the prerequisites are an understanding of probability,...

    What the hell? No wonder the internet slows down for no apparent reason:

    User: What are the chances of my packet getting through today?

    Network admin: Eh, about 83% today. Maybe your odds will improve tomorrow.

Our OS who art in CPU, UNIX be thy name. Thy programs run, thy syscalls done, In kernel as it is in user!

Working...