Software Defined Networking -- changes everything?

mguy

So the industry is a buzz with SDN right now with acquisitions left and right by major companies

Let me google that for you

Basically the networking gear will be given "cloud" like properties and this new technology challenges the status-quo. It could end Cisco, replaced by something else.

I'm not an expert, but as a guy who sees CCNP/CCIE path as a possibility, major technological innovations like this makes me hold off. CCIE study with 1500 hours doesn't look as appealing anymore now that there are new major players, new paradigms, and Cisco dominance is threatened.

How will this affect networking jobs? Thoughts?

kurosaki00

lol cisco dominance threatened?

I'm sure at that "cloud" there are invisible/intangible stuff running that dont need a DBA, a network engineer or a sys eng

Same thing has been going on with Sys Admin and cloud
and I havent seen any reduced demand for dbas or sysadmins
more like they just integrate "cloud technologies" to the position duties

mguy

Actually at the moment, nothing is certain. Nothing is certain in IT. We make bets that'll hopefully payoff.

kurosaki00

The one actitude you shouldnt take is the "Hold off" in your studies
never.
You should always keep learning stuff and keeping up to date
if you do that, it doesnt matter if suddenly red is the new black
because as a Pro IT guy/gal you will know about red months before it goes into full force

mguy

Not necessarily. You should also be business smart about your choice of which technology to study given information available to you.

Would you learn Windows XP just to learn it when the trend is to move away from it? or instead learn the new technology/platform?

A good IT professional can quickly reverse their directions and head towards a new future.

discount81

You laugh at the phrase Ciscos dominance threatened, and while I agree right now Cisco is still the most dominant name in networking, I have to say this.
You probably were not in the industry when Nortel was around, not that long ago Nortel was the most dominant player in the Telco industry.
Now it is irrelevant, some people probably don't even realize it still exists.

Things change fast, especially in the world of technology, I am not an expert in Network/WAN but I feel Juniper has a better product now and sometimes that is all it takes is a company falling behind a little bit.

Roguetadhg

Oh, you mean like Open Flow?

kurosaki00

mguy wrote: »

Not necessarily. You should also be business smart about your choice of which technology to study given information available to you.

Would you learn Windows XP just to learn it when the trend is to move away from it? or instead learn the new technology/platform?

A good IT professional can quickly reverse their directions and head towards a new future.

Thats a bad example
because Win XP is not a current technology


I didnt said anything about the logical fact of inform yourself of what do you want to do and why before going for it
but maybe some people dont see that as logical?
Its obvious you have to be sure of what you study for, there are many variables, money, family, interest

About the nortel comment
I still see nortel A LOT
Mostly gear but bb runs other technology
But guess what, what you learned with nortel, about how network works, how to deploy stuff
still applies!
And guess what? When this new technology comes out, YOU have the edge from people who didnt worked with the "older" stuff.

You guys need to chill
Even if in 6 months Cisco went to 2nd place
MCSE 2k3 is still very relevant nowadays.

alxx

problem is security as sdn opens up more holes and new ways for hackers to screw with things.

Someone still has to make the hardware though.

Some of it seems to be a continuation of the dev ops movement from google and others - have one set of people doing development and operations instead of two separate sets (or more) doing development , testing and implementation.
It might cut costs but there are quite big risks.

If it helps bring a bit more standardisation of equipment and interfaces , it could be a good thing

discount81

kurosaki00 wrote: »

About the nortel comment
I still see nortel A LOT
Mostly gear but bb runs other technology
But guess what, what you learned with nortel, about how network works, how to deploy stuff
still applies!
And guess what? When this new technology comes out, YOU have the edge from people who didn't worked with the "older" stuff.

Yeah there is still a lot of Nortel PBX systems around, a lot of companies probably hard to find the cost justification for switching to new technology PBX systems.
I had 5 vendors come in to give a presentation of changing the Nortel PBX at one of our sites, and the only one who could come up with an ROI under 5 years was offering a Toshiba system, so I am still on the fence if it is even worth changing when the Nortel system still works and most of the people at that site are dinosaurs and would have absolutely no use for these extra features they would get.

I do agree with all your points, all I was saying is I wouldn't say it is completely impossible for Cisco to lose it's dominance as it has happened before.

chrisone

mguy wrote: »

I'm not an expert, but as a guy who sees CCNP/CCIE path as a possibility, major technological innovations like this makes me hold off.

How will this affect networking jobs? Thoughts?

As a guy who has been into IT for more than 12 years, No, cisco is not going away, heck cisco has "threatened" other major companies in the server , wireless, and voice markets. You honestly think they wouldn't have something to offer if in fact the technology changed? You underestimate the values, goals, and responsibilities of CEO's. To think a major technology company would just sit back and let technology run away from them.

ptilsen

Cisco will jump into SDN and probably excel. Cisco isn't going away.

Will SDN really replace traditional network infrastructure? Very unlikely. Will it supplant a huge part of that infrastructure? Very likely. Will it mean the end of the Cisco certification lines' de facto collective status as the preferred network implementation certifications? Very unlikely.

Will CCNPs and CCIEs be lining up for food stamps and unemployment? Absolutely not. SDN will not eliminate the need for expert networking professionals, and right now there's no reason to believe it will even reduce that need. If anything, I see it as part of a larger trend towards requiring greater expertise at the expense of a reduced need in inexpert professionals. In other words, SDN means , if anything, CCNA doesn't get you as far as it used to, but CCNP gets you further.

astorrs

ptilsen wrote: »

Cisco will jump into SDN and probably excel. Cisco isn't going away.

They're making heavy investments, have a look at Cisco ONE (Open Network Environment):

Cisco Open Network Environment - Cisco Systems

And while some of them are defensive in nature, they're still pretty cool: Cisco?s One Platform Kit (onePK)* [Networking Software (IOS & NX-OS)] - Cisco Systems

networker050184

Technology is always changing. That shouldn't stop you from pursuing your goals. You don't think guys with expert level networking knowledge are going to be running these SDN networks?

eansdad

Cisco threatened ... Umm No ... They just bought Meraki, Cloupia and Cariden Tech. If anything they will pioneer it (or buy the company that does).

Now the biggest issues I see with are
1. What happens to your network if the service/server is interupted? Do the dumb switches failover into hubs or do they shut down? 2. How much more traffic does this put onto the network? or Will this be set up on a secondary network for just infrastructure?
3. Cost, buying more equipment to manage the equipment that used to do the job itself sounds like a waste unless you can save a good amount of money.
4. Security/Reliabilty - Something about my switches/routers possibly needing M$ OS in order to run worries me.
5. Is this going to be a new form of IaaS? Having a hosted provider or managed solution for my network seems more like a head ache then just hiring someone to do it on site.

I don't see this taking off as fast as other "cloud" solutions have.

networker050184

I'll take a shot at a few of these from what I know about the technology.

eansdad wrote: »

1. What happens to your network if the service/server is interupted? Do the dumb switches failover into hubs or do they shut down?

It's like basic router architecture these days with an SUP/RE passing info down to line cards. A copy of the forwarding table is stored on the local device and can continue to function even if the 'brains' are missing. It just will not get updated forwarding information.

eansdad wrote: »

2. How much more traffic does this put onto the network? or Will this be set up on a secondary network for just infrastructure?

No clue on this one.

eansdad wrote: »

3. Cost, buying more equipment to manage the equipment that used to do the job itself sounds like a waste unless you can save a good amount of money.

The 'dumb' switches should be drastically cheaper. They will basically just be forwarding planes, or extended line cards. The real money s spent on the intelligence.

eansdad wrote: »

4. Security/Reliabilty - Something about my switches/routers possibly needing M$ OS in order to run worries me.

I doubt they will run on a MS server. I'd assume most will be appliance based with a locked down nix flavor of OS that won't be interacted with much outside of TAC.

eansdad wrote: »

5. Is this going to be a new form of IaaS? Having a hosted provider or managed solution for my network seems more like a head ache then just hiring someone to do it on site.

There are plenty of outsourced networks now. I don't think this will change much if anything about that model. Some shops prefer to admin from in-house some don't.