CCIE longevity??

sea_turtle

dppagc wrote: »

You know I was thinking, what would Brian be doing if CCIE is obsolete?

maybe he would finally update the SP track materials.....

Iristheangel

sea_turtle wrote: »

maybe he would finally update the SP track materials.....

Or maybe he'll go on a cruise without trying to turn it into a paid vacation

d4nz1g

networker050184 wrote: »

I don't think anyone even knows what SDN means these days. It's used to describe so many different things it''s as useless as the term cloud.

That said it all

razar

Thanks for the input everyone, my aim is to continue and take the written by the end of this year and lab next year. Even if the technologies have moved on I still think that it will be valuable at least in the short-medium term.

Just wanted to gauge the feeling and opinions from the rest of the community as it's something that has been in the back of my head for a while.

For me the main thing is that since I've started studying again, I feel like I'm learning a lot and it's making me better at my job right now. And hopefully will put me in good shape for the future.

bender000

I'll just add three more points here

Scripting/automation/programming is useless if you don't know what you're scripting/automating/programming. Can I use Ansible to throw up an Arista deployment? sure. Can I automate the inventory gathering via python dynamic inventory? sure. Can I do it if I only learn ansible and do not learn VXLAN or understand a L3 routed ECMP underlay/anycast IP gw topology or the various routing designs for north/south egress? not a snowball's chance in hell. What about BGP only leaf/spine (same AS leaf layer or different?) vs OSPF leaf/spine (that crazy Cumulus design with ip unnumbered everywhere)? Do you need a CCIE to understand what I just said? not really, but it helps, and the more stripes you have, the more fun gigs you get. And guess what, the larger deployments I've seen are buying into cloudvision because guess what, the customers don't have scripting engineers who will whip up their own solution via pyEAPI and ansible, they want a vendor GUI/system.... back to clicking buttons....

Secondly, this is all sounding a lot like the sky is falling scenario when virtualisation first became common. Oh noes they'll take all our sysadmin jerbs. Guess what happened? the number of VMs exploded and the complexity of the VM infrastructure exploded, more admins than ever. Same thing will happen with automation and SDN: it lets us do more, so guess what, we'll go and do more, and the complexity will keep increasing, which will require more and more engineers. Even right now, its impossible to get a handle on all the emerging DC based SDN solutions - and that's just thinking about the 'big 3' (openstack, ACI, NSX), then there's all the underlay/overlay standards and solutions, then there's contrail, all the various openflow implementations..... then traditional carrier MPLS orchestration with BGP-LS and PCEP.... anyone who reckons they know more than 1 or 2 well is lying through their teeth because to get that knowledge they would have spent the last 2-3 years working on a big project, and guess how long all these solutions have been around for.

Finally, what do you think all these DCI SDN blocks are using for north south/edge transit.... and what do you think all the various 'SDN domains' will use for connecting to each other / the outside world..... knowing BGP well is not going to go out of style in a hurry.

fredrikjj

bender000 wrote: »

Scripting/automation/programming is useless if you don't know what you're scripting/automating/programming. Can I use Ansible to throw up an Arista deployment? sure. Can I automate the inventory gathering via python dynamic inventory? sure. Can I do it if I only learn ansible and do not learn VXLAN or understand a L3 routed ECMP underlay/anycast IP gw topology or the various routing designs for north/south egress?

I don't think anyone is actually predicting that knowledge of the current control plane will become obsolete any time soon. However, that is not a strong argument for or against studying for a CCIE in my opinion because the ROI is decided by demand for the skill set, not what technology is actually used. Let's say that 4 network engineers are replaced by 1 neteng and 3 devs where the neteng supplies the networking skills and the 3 devs work on implementing the automation aspects. The networking skill set is still highly relevant in that scenario, but demand is suddenly much less because you need fewer people.

Also, even if knowledge of some tech is important to a business, that does not mean that people making decisions will automatically behave rationally and hire people with that knowledge. Trends can matter, and if people are convinced that coding is the most important thing, maybe they will hire that software developer that's "interested in networking" instead of the network engineer that has dedicated many years to that specialization. Perhaps their network will suck because of it, but that might not even matter or be obvious until some failure happens two years later, or whatever.

So in short, technology is not what you should primarily look at when deciding the ROI on some education. Empirical data on hiring decisions is much more valuable.

joelsfood

Let the developers run your network for a few months, and you'll be back begging for people who actually understand networking, no matter how good the automation is

networker050184

Well the key is the people that are your network developers know networking. You wouldn't throw developers without network skills at it alone. That'd be silly. The people that do the developing for my network we are all high level network architects, CCIEs etc. as well. Tell some employers you know how to automate tasks these days and you'l be golden. Really sets you apart.

Bardlebee

Even in the largest organizations, the most organized teams and the most dynamic places have a hodge podge of different networks going on. Through acquisitions and all that, I don't see demand decreasing for the CCIE, I don't agree to the point that you will have 1 engineer and 3 devs as opposed to 4 engineers.

I think at worst you may have 1 engineer, 2 engineer/dev hybrids (leaning heavily on engineering) and 1 pure dev with networking background.

Networks, for the foreseeable future are too dynamic in my mind to automate enough for the CCIE to be out of style. And as someone else mentioned, the CCIE is always updated, so its hard to not stay relevant.

I think the main question here is, is it worth a year of your life and time to study for it. And this question could be asked of anything! You could certainly ask yourself, at a CCNA or low CCNP level knowledge, would doing dev ops for a year help your career more then doing something else for a year. These are things you have to weigh. I don't think there is more evidence for either one to be honest with you. I think right now and for the next decade I think both will be needed almost equally honestly. And you will never not find a job for either as well.

networker050184

Bardlebee wrote: »

I think the main question here is, is it worth a year of your life and time to study for it.

Indeed that is the question. The question isn't whether routing and switching skills will be needed in the future that most seem to be answering. Of course we'll still need to know ethernet, routing, etc. For someone looking to stay on the cutting edge of our field and already has a high level knowledge in networking I think there are more important things career wise to concentrate on.

Bardlebee

networker050184 wrote: »

Indeed that is the question. The question isn't whether routing and switching skills will be needed in the future that most seem to be answering. Of course we'll still need to know ethernet, routing, etc. For someone looking to stay on the cutting edge of our field and already has a high level knowledge in networking I think there are more important things career wise to concentrate on.

The argument could and maybe should be made that is it worth focusing on ONE THING for a full year to be a jedi master in it. And I would say that, yes it is. If you focus on pure dev for a year, never touch SP/DC/Wireless or whatever, you're going to be a badass in that art. And there are PLENTY of places looking for badasses in that art. But I feel the same goes for the CCIE, plenty of people are looking for that depth, I believe.

More importantly we know that networks are here. We don't KNOW dev work will "take our jobs" and nor have we seen that happen yet. I think its important to use automation as a tool and not a job replacer, as I don't think it can be. But if I'm proved wrong, great. I'll be studying scripting/automation after my CCIE.

If you start the CCIE now, there is not a snowball chance in hell that your job will be automated next year, I believe we've still got a good 10 years before that concern should be coming up, maybe worst case 5.

I mean, a lot of places want generalists and most places have generalists. And I would agree that there are less jobs for CCIE's specifically, but as I pointed out, the CCIE is more of an intangible benefit of finishing something of high prestige that employers can look at, not just now, but when our jobs are obsolete (won't happen anytime soon).

I still tell people when coming into this field now "Learn scripting" because I want to prepare them to be the best. And I think its important for a great engineer to know scripting/automation. But I don't think it takes away from the CCIE, rather I think it adds to the whole picture.

This is a bit long winded, but at the end of the day, who do I want to plan out a DC consolidation between multiple locations around the globe? Probably a CCIE. Who do I want to automate and standardize post-consolidation, probably a dev ops guy, in conjunction with a networking guy. I really cannot see a pure dev guy, with no networking knowledge being able to do any of that at all. It would have to be someone with a CCNP at least in knowledge/practice. And if they are CCIE + scripting. You can write your own check I suppose.

networker050184

Of course people should still spend time to become experts in networking technologies. That's mutually exclusive to spending time and money on a CCIE cert though.

silver145

the usual "is it worth it" "when will it be automated "when will the stock market crash"


Every year some one on a forum has wrote "when will it be over, is it still worth it" even during the tech bubble 2000's it was all the same. Get it while you can, few years it will all be automated.....still kicking

gorebrush

Agreed - while some companies will embrace it, the vast majority wont rush. Going to be an evolutionary technology for sure.

bender000

networker050184 wrote: »

Of course people should still spend time to become experts in networking technologies. That's mutually exclusive to spending time and money on a CCIE cert though.

Disagree strongly (sort of depending on which CCIE). The meat and potatoes of a CCIE RS is routing protocols and the foundations for MPLS-VPN. That stuff is not going away in a huge rush. Moreover, the more you dig into all the overlay/underlay topologies, the more everything looks like variants of MPLS logic (esp now with MP-BGP control planes coming back in vogue with EVPN for example).

I'm strongly vendor agnostic and will recommend palos/junipers/riverbeds etc. to my clients any day, I will bag ACI and push Arista and/or NSX all day long, but the CCIE RS is still the heavyweight of the field, and justifiably so.

Also in terms of 'paper' usefulness, its the king of the hill by a LONG shot. Its a guaranteed interview and a 10% higher asking price for any job right out of the gate. Heck its so well known that non-technical people don't even click that there are separate streams so you get tagged as the expert in all things networky, even if your digits are in RS but the topic is for example wireless or whatever.... (and often quite scary.... WTF do I know about OTV... oh well, here we go again... LOL).

Finally, I don't know why its always framed as an either-or scenario. Sure during the last 3-4 months you're going to be head down bum up at your labs, but outside of that nothing makes it an exclusive thing.

To be brutally honest, any argument I hear that claims its not worth doing an IE, just sounds like a combination of excuses and sour grapes.

networker050184

silver145 wrote: »

Every year some one on a forum has wrote "when will it be over, is it still worth it" even during the tech bubble 2000's it was all the same. Get it while you can, few years it will all be automated.....still kicking

Indeed and every year I have always answered yes it is still worth it. I've always been a big proponent of certification. I just see the tides shifting lately on this one in particular.

And again I'm not saying any of these particular skills will not be useful. I'm just starting to see the field lean more towards other skills being the most sought after return on investment of time wise.

Put it this way, if someone had one year to to really set themselves up for their future career wise I'd definitely suggest them to get a CCNP and some scripting/automation skills rather than simply spending that year on a CCIE. I think we're finally to that point in our industry.

sea_turtle

bender000 wrote: »

To be brutally honest, any argument I hear that claims its not worth doing an IE, just sounds like a combination of excuses and sour grapes.

100% correct.

the other thing i keep seeing is that the demand will lower for pure network players as the supply will stay the same or perhaps increase....

let me put this out to everyone here, and be honest about yourself and your co-workers. how many quality network people do you have? how many that are always learning and evolving? how many can you trust to do the job correct so you do not get called at night??

then on the flip side how many are you pretty sure are brain dumpers? or got the CCNA/CCNP a few years ago and just come to work every day not really caring to advance?

the demand will likely decrease surely but the supply of quality engineers has always been on the lower side pre SDN (similar to the transport world as its treated more like a traditional apprentice career path). with SDN further becoming a reality from a buzzword that means the focus will shift from non stop network bootcamps to non stop python bootcamps helping to lower the supply even further.

lostindaylight

sea_turtle wrote: »

100% correct.

let me put this out to everyone here, and be honest about yourself and your co-workers. how many quality network people do you have? how many that are always learning and evolving? how many can you trust to do the job correct so you do not get called at night??

then on the flip side how many are you pretty sure are brain dumpers? or got the CCNA/CCNP a few years ago and just come to work every day not really caring to advance?

Hi Sea Turtle.

I work with guys like that yes.

My answer has a small twist though.

At least in a smaller shop one driven and motivated engineer can raise the level across the board. By bringing new technologies and designs into play and constantly pressing forward it forces the issue: i.e. they have to respond and do some learning, or be exposed and have management questioning their commitment to being a professional. And there's a personal pride aspect of it. No one like being shown up day after day.

A CCIE type of engineer can raise the level of everyone around them. You can tilt the field and change things for the better! That's a benefit I don't see people talking about too much, the discussion generally revolves around the struggle and benefits of the individual.

Legacy User

If its worth it? Hmm well do you think it will benefit you in your current job or will you be looking for other job prospects once you do get it? CCIE level knowledge alone is priceless and will make you an asset wherever you go. While its very possible in 5 years everything may be gui based but as of right now there are plenty of places that work on the cli. Having that knowledge will benefit you now and noone can predict what technology will be mainstream in 5 years but when that time comes you just have to adapt if you want to remain relevant. But regardless of how its implemented cli or gui routing protocols, vlans, trunking, etc will still follow the same principles so they will still need someone that understands how the underlying technology works.

At my place I haven't heard any talk about apush for certifications but a push to learn the network in detail which is allCisco RS, Wireless, DC at least thats whats required to learn on my team, and theyfollow cisco best practices for everything. So naturally I am going to get anycertifications relevant to help me become better at my job and what can get me more projects in acertain technology and CCIE level knowledge would only benefit me so in my case its definitely worth it.

networker050184

I don't get the whole GUI thing. I don't think that has anything to do with the conversation at hand, but seems to be what most are harping on. You can have SDN, automation, any of that stuff without a GUI.

Legacy User

I was basing everything from the OP he said he's working with alot of stuff on GUI at his place of work. Personally I don't really get SDN since it still seems pretty obscure. Is python scripting automation still considered SDN? My understanding from cisco standpoint how it was explained as a window that you can select the vlans to change or whatever to streamline operations, that sounded like a gui to me.

networker050184

No automation is not SDN. You can certainly automate a "regular" network as well. I use ansible mostly as well as some home brew python applications. My ansible deployment is 100% command line. None of that is SDN. SDN is basically a centralizing control plane. Most systems tend to have a GUI front end available, but just like an ASA and ASDM you can still do everything from the CLI if that is what you prefer.

joelsfood

"Andrew Clay Shafer ‏@littleidea 15h15 hours ago 7th rule of devops: no one should work on automating anything they couldn’t operate by hand"

Don't expect any GUI/SDN/Devops/etc to replace Network engineers (or system engineers, storage engineers, etc). If you want your CCIE, go get it. You will always find people to discourage you, and only internal motivation will get you through it. So threads like this, while interesting, will never be enough to motivate someone who doesn't really have the passion for a CCIE (or VCIX, etc), and shouldn't be enough to discourage the interest.

Chinook

Have you given though to acquiring multiple CCNP's (R&S, Security, Voice, Wireless) versus just a CCIE. Gives you a broader range of skills & they're less likely to become null & void.

razar

Chinook wrote: »

Have you given though to acquiring multiple CCNP's (R&S, Security, Voice, Wireless) versus just a CCIE. Gives you a broader range of skills & they're less likely to become null & void.

Yes I have thought about that actually, I'm interested in design and DC and like the new materials in the new CCDA in particular. I may read the Cisco Press book for that when it comes out and take that exam. My CCNP expires at the end of this year so I'd like to take the CCIE Written before then, my study has been going well recently and if I keep on track I think I'd be able to achieve that. Ideally I'd like to get a ccie in R&S and then move onto learning some of the newer technologies in more detail. Or work towards CCDE.