February 12, 2025 • 48 mins
The (PA)NUG podcast panel of William Collins, Andy Lapteff, Ned Bellavance, and Drew Conry-Murray highlights the evolving landscape of network engineering, focusing on the essential skills required for future success, including automation and cloud technologies. Panelists emphasize the importance of foundational networking knowledge, mentorship, and the integration of security into network design, while also addressing the challenges and opportunities presented by technological advancements.
• Exploration of the changing network engineering landscape
• Emphasis on the importance of automation and scripting skills
• Discussion on retaining fundamental knowledge in networking
• Insights on securing networks within cloud infrastructure
• Importance of mentorship and continuous learning
• Perspectives on the future of AI in network engineering
Join the (US)NUA today:
https://www.usnua.com/membership
Find everything AONE right here: https://linktr.ee/artofneteng
Mark as Played
Transcript
Episode Transcript
Available transcripts are automatically generated. Complete accuracy is not guaranteed.
Speaker 1 (00:01):
Hey folks, back in October 2024, I had the pleasure
of participating in USNUA'sfirst podcast panel, where Drew
Conroy-Murray, ned Belovance,william Collins and I discussed
the current state of the networkengineering profession and what
it takes to create and sustaina successful career in our
ever-evolving field.
The fine folks at USNUA haveallowed the art of network
(00:24):
engineering to share the showwith you on our platform and I
think you'll enjoy the dynamicconversation that follows, thank
you.
They host networking usergroups, or NUGS, all over the
country, where free food, freebeer and free information is
(00:46):
presented in a relaxedatmosphere.
For more information on theUSNUA, as well as access to a
variety of member-onlyeducational resources, head over
to usnuacom and check out themembership section to become a
member of this fantastic groupof network engineering
professionals today.
Now sit back.
Enjoy our PA Nugs podcast panelfrom October 2024.
(01:09):
As always, thanks for listening.
Speaker 2 (01:16):
I think I mentioned when we started this thing.
This is the first time we'redoing something like this.
It's a joint initiative betweensome different people from the
community and the USNUA toreally bring together what we're
referring to as the podcastpanel, right?
Just basically people that areinvolved in the community,
people who are in the know andpeople who have a pulse on the
industry, and what we're tryingto do is really get their
(01:38):
perspective, you know,essentially put a face to the
voices that everybody'slistening to on these podcasts.
Right, and I'll let everybodyback here behind me give an
introduction for themselves.
So I don't butcher it from myside of things.
But the one thing I did want tomention as we get into this is
if you ask a question, I'mprobably going to repeat it.
They mic'd me up for somereason so that way they can
record everything.
I don't know podcasters, weirdpeople.
(02:00):
I also have a second mic.
If you have a comment, chancesare I'm going to pass you a mic.
So if there's a comment, feelfree to, like you know, wave
your hand, throw something at me, whatever's easiest, and if
it's a question, I'll probablyjust repeat it so that way they
can respond to it.
I passed around a couple ofquestion sheets.
I guess we'll call them.
If there's something on thereyou really want to talk about,
you know, feel free toacknowledge it.
(02:21):
Right, you know, we'll getthere in some way, shape or form
.
I'm probably going to startwith changing landscape and
we'll go from there, but we'lldive in.
But, eric, if you want to go tothe next one for me, there we
go.
I'm going to go left to right,although the pictures don't line
up that way, and I'm going tolet everybody introduce
themselves here real quick andwe'll dive in big enterprise.
Speaker 3 (02:52):
Started out, my first job was actually managing CSS,
load balancers, certificates,which was like the worst job
ever.
I would have rather worked fastfood at the time, but you know
I put in the time.
So I started in the data centerand then pivoted to cloud.
When cloud started reallybecoming a thing, did the
multi-cloud thing, solved a lotof problems in the enterprise.
And then now I work for astartup for three years my first
(03:14):
time working for a vendor andthen I started the cloud gambit
podcast and I co-host withYvonne Sharp and really what we
do is we focus on a lot of theadjacent areas to network
engineering.
So my core was networkengineering.
But one of the cheat codes I hadwhen I was in enterprise I'm
curious, I like talking topeople.
(03:36):
So that was how I was able toget things over the finish line
in the enterprise.
Hey, I'm gonna go to talk tosecurity, I'm gonna talk to the
developers, and it's some ofthose points like we couldn't
work from home.
This is a while back.
And so, hey, they're in theoffice.
I just go over and talk to them.
You know they're right there.
So I figured, hey, why not dothat with some of these engaging
conversations.
(03:56):
So we talked to everybody from,like network engineers,
founders of startups, venturecapitalists, because it all,
like, really is closely meshedtogether, so happy to be here.
Speaker 1 (04:07):
Hey guys, andy Laptev , co-founder, co-host, art of
Network Engineering.
I started out in tech workingat a Verizon Central office
pulling wires for POTS lines andDSLs and getting the bejesus
shocked out of me when you wouldtouch the actual things.
Cable guy Comcast for fiveyears worked my way into their
NOC with a CCNA and I got a jobin fintech for six years
(04:30):
managing their global WANs.
Burnout there and right towardthe end of my burnout got
plucked at Juniper to do someproduct management stuff, some
customer experience work andwhat else can I tell you?
Our show, I guess, because thisis a podcast panel, we wanted
to tell the stories of networkengineers doing the job.
So people like you, theoperators we didn't really think
(04:53):
that audience was all that wellserved at the time.
So you know what's it likeworking in an Ock, what's it
like climbing that ladder,what's it like being a network
engineer, and it's been awesome.
I've gotten to meet a ton ofpeople.
I mean we were just at CiscoLive.
I went to my first Cisco Live.
I couldn't walk 20 feet withoutsomebody being like oh my God,
are you the guy from the thing?
I'm like whoa.
So it's just been beyond mywildest dreams.
(05:13):
Started in COVID.
It wasn't like a plan, we werejust home, we were studying for
start a podcast.
These conversations in ourstudy group are kind of funny
and it just kind of took off.
But great to be here.
Thanks everybody for coming.
Thanks Andy.
Speaker 5 (05:28):
Hi, I'm Ned Belovance and let's see, we're doing
history here.
So it all started on a warmsummer day in 1970.
No, so I got my start in techon Helpdesk as I think a lot of
people might and worked my wayup through there to be a desktop
(05:49):
engineer.
And then I started getting intonetworking, took my CCNA, tried
to go for my CCNP, failed thefirst time, said forget this
stuff, I'm going to learn VMware, and never looked back.
So became a systemsadministrator and then a
consultant, and I worked inconsulting for eight years and
while I was doing that, cloudhappened.
(06:09):
So I had to start learning AWSand Azure, because that's what
our clients wanted us to setstuff up on.
And hey, networking is there.
So now I have to remember allthe stuff I learned in CCNA and
also then augment that withadditional things that were
unique or somewhat unique to thecloud.
So that's what I was doing foreight years, and then I decided
(06:32):
to get out of the consulting bizbecause that is a lot of stress
Stress, I think, is the rightword I'm looking for and so I
went into full-time contentcreation and being a technical
educator.
So I have a bunch of courses onPluralsight around things like
Terraform, vault, azure, and Ialso teach live stuff.
And, oh, I have a podcastbecause this is the podcast
(06:54):
panel.
I have two podcasts because oneis never enough.
Oh God, they're like Fritos.
So, seriously, I'm on like mysixth podcast at this point.
So one is called Day Two DevOps.
That is on the Packet PushersNetwork that I host with Kyler
Middleton.
She's awesome.
We talk about anything looselyrelated to DevOps.
(07:15):
So that could be careeradvancement, it could be a new
tool, could be just someonewho's got a cool thing they're
doing at work and they want totalk about it.
And I also host a podcast calledChaos Lever with my buddy,
chris, where we talk about thehistory of technology.
So we recently did a podcast.
We recently covered that wholeWordPress thing that's happening
(07:36):
.
I don't know how many peopleare aware of the nonsense that's
happening with WordPress.
I don't really care about that,but it gave me a chance to talk
about BBSs and blogging and howall of that morphed into the
fact that WordPress runs like50% of websites now.
So if you like tech history alittle bit, should I give them a
little preview?
Go for it.
I already forgot what I'mworking on for next week.
(07:58):
That's terrible.
I spent all day writing no, I'mkidding, excel is turning 40.
Microsoft Excel is turning 40.
Microsoft Excel is turning 40.
And it also runs like 70% offinance in the world.
So I thought it'd be fun tolook back at spreadsheets a
little bit.
Also, like 90% of automation,most networking is really just
(08:19):
configured through a spreadsheet.
So I thought it'd beinteresting to look back at the
past of spreadsheets where theycame from and how they evolved.
Speaker 6 (08:27):
And that's me.
I'm Drew Connery Murray.
I am kind of the odd person outhere.
I came into this industry onthe reporting side, so I've been
covering the tech industry as areporter, writer, editor,
blogger, podcaster for almost 30years now 20, 25 years now.
I've worked for outfits likeNetwork Magazine, network
(08:52):
Computing Information Week.
For a while, when the Interopconference was running, I was
the director.
I was responsible for bringingin all the technical
presentations that happened atInterop.
Through my role as an editor atone of the magazines I was
working for, I got to know EthanBanks and Greg Farrow, who
(09:14):
founded the Packet PushersNetwork.
Greg invited me to start apodcast with him called Network
Break, which is like a weekly ITnews show.
We started it.
We were having a good time.
Then I got laid off fromInterop.
So I went to Greg and was likeif you want to find another
co-host, you know somebodyconnected in the industry,
that's fine.
He's like do you want a job?
I was like, well, yeah, I needa job.
That's how I joined PacketPushers.
(09:34):
I've been there many years nowand I co-host three podcasts
Network Break.
I co-host Heavy Networking,which is sort of the flagship
networking podcast, and a recentpodcast that I started with
Jennifer Minella called PacketProtector, which sits at the
intersection of networking andsecurity.
Ned is on the Packet PushersNetwork and Scott Rabban has
Total Network Operations.
(09:55):
We just launched a new podcastcalled Technically Leadership.
If you're looking to get intotechnical leadership, that's a
great show and we've got somereal niche ones like IPv6 Buzz
If you're trying to get yourhands around IPv6, some really
deep content there.
Speaker 2 (10:09):
Given your perspectives right and the
history that all of youmentioned here during the
introductions, the landscape innetwork engineering today right.
Maybe we just start therePerspective of where you've seen
it come from, where it's goingand what you think the next
generations are going to looklike.
For today's network engineers,what they need to know, where
the industry's heading.
Speaker 3 (10:29):
Yes, I remember when I first started doing any sort
of automation it was a long timeago Bash, perl, expect, back to
the hating CSS load balancersthat was what I was using to
solve that problem.
I wanted to automate it and itfreed up a lot of time.
It allowed me to work on otherthings and back then it was a
(10:49):
combination of some differentscripts and we basically had a
jump server and in that jumpserver we gave the developers
the ability to fill in a certainspreadsheet.
They would put it in adirectory and then I had a cron
job run at strategic timesduring pre-approved change
windows that would go out and dothings.
When you look at that and youlook at, okay, it's Linux, it's
(11:10):
a cron job, you're gluing thingstogether.
And then I look at recently,these cloud pipelines I've built
.
It's so eerily similar becauseit's okay, I'm still using Linux
, I'm still gluing thingstogether, I still have this,
this instead of a long-livedserver.
It's ephemeral, like it'scoming up during the pipeline
and it's getting everythingstaged and then it's, you know,
(11:31):
tearing back down.
But there's so manysimilarities there and just
thinking about how much time haspassed and looking at okay,
that was then.
Okay, it was pearl and bash,but now, okay, it's still bash,
and maybe it's Terraform andPython, or Terraform and Golang,
and you know a YAML file thatsits in a Git repo.
It's very similar.
So, yeah, things have changed.
(11:53):
And I think the thing thatreally has changed is you have
networking under the hoodeverywhere.
So even if you look at you knowyou have physical, you go up to
layer three and layer four andthen you go up to the
application networking.
But even as you layer on moreabstractions, under the hood
(12:15):
it's a lot of tunnels, a lot ofreverse proxies and a lot of
things.
But the thing that's actuallychanged and one thing that I
find a lot of network engineersare, I don't want to say, scared
of, but they just want to stayaway from it is the tools that
we use to interact with thesethings.
So, instead of going in andhammering on various CLIs, the
workflows change.
The way that we interact withthese things has changed and
architecture's changed too.
(12:35):
So with the more abstractionlike the other day, I helped
someone that.
So it was a Kubernetes clusterin AWS add an EKS cluster and
they made a change to someTerraform.
I think they changed theversion or something and that
actually tore the clustercompletely down and brought it
back up.
They caused an outage, like theimmutable aspect of this.
(12:56):
But there's a lot of things youcan do that will actually make
configuration changes to thatcluster in place, like maybe a
node group or tags or thingslike that, where you're not
gonna cause an outage.
So really in the architectureworld in the cloud, it's a lot
of reading, it's it is kind ofdry, but you just have to know
how these resources work andwhat their behavior is you
(13:17):
embraced automation way beforeme, or anybody else I know did
so.
Speaker 1 (13:22):
I got my CCNANA in 2012, I think, and there was no
automation on the exam.
There was none being taught.
I got a job at a NOC there wasno automation at all, and five
of six and a half years at theFinTech role there was no
automation.
To speak of what's changed,like to answer your question
what's changed?
Automation and cloud were thetwo big 800 pound gorillas that
(13:43):
kept peeking around the cornerand I kept going get the hell
away from me, I want nothing todo with you.
I am the CLI jockey, right,like that's the job.
And I was brought up maybe in adifferent time.
That was the whole program.
I went to Netacad Like that wasthe job.
So fast forward, years later Ifind myself.
I fought all the changinglandscape as long as I could and
then I thought if I could runaway to a vendor and just smile
(14:08):
and be charming and helpcustomers be happier, like, oh,
this is great, I don't have tostudy anymore and screw that
automation stuff.
And then I got laid off inDecember and I started looking
at job postings.
I'm like shit, they all wantautomation.
You have to know Python orAnsible Infrast.
All want automation.
You have to know Python orAnsible Infrastructure is code.
You have to be versed in acloud.
(14:28):
So really, for me it was onlyout of.
I was beaten in a submission iswhat I'll tell you.
The industry has changed.
If you look at job postingstoday as opposed to even five
years ago, I have a lot ofpeople come to me Because we do
a lot of work in the communityhelping people, and I probably
have seven people right now I'mtrying to help with resumes or
trying to get into the industry.
What are they doing?
And what I told people five orsix years ago get your CCNA, do
(14:52):
some content creation so you canshow your soft skills, your
communication skills, and builda home lab.
It doesn't have to be physical,but you're not going to have
experience and people are goingto call you out in interviews
and a home lab can show you haveexperience.
We don't have any productionexperience.
Well, the hell, I don't.
I got four 2911s, I got 3750sWhatever right and I got a
server.
So it's a way that I got aroundthe no experience thing.
(15:15):
So I say all that to say thatthe past I don't know.
Six or seven months I've beenlearning Python.
I've been learning GitHub.
I haven't gotten into cloud yet, but that's kind of next on my
radar.
I'm reading an LLM book rightnow that Phil Gervasi just
turned me on to, so I'm tryingto get caught up in all the
things I avoided all those years.
There's nothing likeunemployment to wake you up out
(15:37):
of your cognitive bias andre-change your compass of like,
oh, just because I don't want tolearn automation.
Nobody gives a damn.
You know what I mean.
And if you're at a job that youdon't have to do any of that,
good for you.
I was in a job.
I didn't have to do any of that.
I'll end with this.
The job that I had for sixyears that I was a senior at
Edge ran all the WAN projects.
They wouldn't hire me Pythonknowledge.
(16:06):
So I can't even get the jobtoday that I got before at the
same company.
I was a rock star there.
Just because I'm not Pythonproficient, I'm getting there.
I just ran my first script inNetMeco and I logged into a
router and did a show for inwhatever my just visual studio
code.
And I'm also learning Linux.
Right, these things that, likeit was on Scott's thing earlier
Linux automation, you know allthat stuff.
So it is changing and I'm justbusting my ass now trying to
(16:28):
catch up.
Speaker 5 (16:30):
I'm the cloud guy, so I've got to say that the
biggest change has been therollout of the cloud.
And it's not just thathyperscalers exist, but they put
a pressure on the way thateverything else is done Because
they're the 800-pound gorilla,they're the 2,000-pound gorillas
in the room.
Else is done because they'rethe 800 pound grill.
They're the 2,000 poundgorillas in the room.
They're the ones who can drivethe design of specific ASICs
that are going into network gear.
(16:50):
They can determine how systemboards are designed for the
compute that they're running.
They have a lot of influence, Iguess is what I'm saying.
And their philosophy iseverything needs to be
programmable, everything needsto have an API and everything
needs to be automation friendly,because there's no way we can
do this at scale If somebody hasto click on next, next, finish.
(17:11):
And you know, I came fromdesktop support and a world of
Microsoft where a lot of stuffwas next, next, finish, and I
also had to learn automationslightly different context, but
same basic thing.
I was managing Exchange 2003back in, you know, 2003.
(17:32):
And then Exchange 2007 came outand, for those who don't know,
exchange 2007 was theintroduction of PowerShell into
the Microsoft ecosystem for real, and there were certain things
you could only do in Exchange2007 through PowerShell.
You couldn't go into the GUIanymore, and that was great for
(17:54):
big organizations who were likewe want to run these scripts at
scale and do things, and itsucked.
When you're an SMB and suddenlyyou are also responsible for
writing PowerShell, you canguess where I was.
But it forced me into the worldof automation and by doing that
I learned PowerShell, which isscripting, programming, call it
whatever we want.
(18:15):
But then I was able to applythat skill to VMware and then I
was able to apply it to Azure.
When that rolled out and then,after realizing that using
PowerShell to control Azure iskind of a pain, I discovered
Terraform and I was like, ohokay, so I can do this
declaratively now.
But the core premise ofTerraform and all these other
tools is the fact that there'sgoing to be an API you can talk
(18:37):
to.
So I guess the biggest changeis the fact that you have to
shift the mentality a little bitaway from the CLI and on to how
do I interact with an API orwrite a tool that knows how to
talk to the API?
And that's been a struggle forme as much as it's been for any
network engineer, but slowlywe're all getting there right?
Speaker 6 (18:57):
Yeah, it's hard to disagree with anything that's
been said.
I think, as Scott mentionedearlier in his presentation,
networking sort of fundamentalshave kind of we're not fighting
about TCPIP versus ATM orwhatever.
We're not battling over tokenring versus Ethernet.
That's been decided already andnow we've got that structure
and we're building on top of itin a data center.
(19:17):
It's LeafSpy and EVP and VXLANand we don't need to argue about
it anymore.
Where things are going is towardautomation, is toward more
software, developer-likeconstructs like GitHub, like
APIs, like being comfortablewith Python, ansible and these
new tooling things.
Cli skills that you put a lotof work into may not be as
relevant, but the thing that isstill relevant is all of that
(19:42):
fundamental information that youhave about how IP works, about
how Ethernet works, aboutsubnetting, about addressing.
That's still essential,regardless of whether you're
doing it in the cloud, on-prem,through some new software
interface or right into the CLI.
So it's not an abandon all hope.
You've got the fundamentalsthat are essential now, like
this gentleman from Oracle said,oracle doesn't have a cloud
(20:02):
business without a network.
Speaker 5 (20:03):
Somebody had mentioned I think it was, I
forget which one of you, and I'msorry.
Staffing is a major concernFinding people that have the
fundamental skills and that whenyou onboard someone, it's more
about a history lesson becausethey don't know that stuff.
Maybe they're an expert inPython but they have no idea how
to set up a trunk for VLANs.
(20:24):
They're like a V, what I'm notfamiliar with the terminology,
and so if you already have thosecore skills around networking,
you understand the fundamentals.
Like programming is hard butit's not that hard and I mean
I've seen people that havelearned programming.
I guarantee you can all learnit.
Um, what's really really hardis the last 10, 15, 20 years
(20:46):
you've put into learning networkfundamentals and that's
something that's irreplaceableand someone can't just
accelerate that in a three-monthcourse and suddenly understand
all the fundamentals.
So you got to leg up on people.
That way it it's just you haveto continue to learn and grow
your skillset to accommodate thetools that are driving that
underlay.
Speaker 2 (21:05):
So I do want to pass the mic around the room here a
little bit as well.
Right, thinking about wherethey've seen things come from to
where they think things aregoing.
Anybody with a strong opinionin the room?
Just relative to some of whatthey mentioned?
It sounds like cloud andautomation, two of the big ones.
Right, I'm sure we can allagree we're seeing some of that
within our roles and what we'redoing now.
(21:26):
But is there anything elseyou're seeing outside of those
two realms that you think ispertinent to what we're talking
about when we look at wherenetwork engineering or
engineering in general is headed?
Speaker 7 (21:34):
So one of the things that I've seen, I've experienced
, is we went through a huge datacenter migration right, and
part of it is networking rightand you're part of that team to
migrate apps from on-prem to nowinto AWS right, and you're kind
of forced into the cloud worldand, like you said, there's
(21:57):
networking component, transitgateways, and you're also, as a
network engineer, expected tounderstand security right, and
so that's a skill that you knowwe haven't touched on tonight,
but you have to know it right,because there's networking in
every firewall that you touchright.
There's BGP, there's OSPF right,and if you're going to do
(22:19):
direct connect with AWS or anyof these other cloud providers,
you have to understand thatstuff right and you have to
understand how you're going to.
You know peer with othernetwork, you know with your AWS
cloud and everything like that.
So I think that's one of thethings that I've experienced is
that traditional networking is agreat base when you have to
(22:41):
learn cloud skills.
Once you get into the cloud andyou're like, oh, these are
things I've done for the last 20years, it's that much easier.
Speaker 2 (22:49):
One of the interesting things about the
networking industry and I'msaying this as not one of the
panelists so cut me off.
If Andy later on edited it out,go ahead.
So cut me off.
If, like Andy later on editedout, go ahead.
But at times you almost have tobecome, because the network is
touching everything right andyou have people that are
specific to individual domains,like you mentioned.
You know you're going to havesecurity, you're going to have
(23:09):
infrastructure, you're going tohave network, so on and so forth
, but you almost have to be insome way, shape or form, kind of
like that jack of all trades.
Speaker 7 (23:15):
You do Network or form, kind of like that jack of
all trades.
You do and you are, and and forlack of better term, you're
sometimes you.
You get asked to join a call,right, and you're the adult in
the room.
You are, you're the adult inthe room because it's seven
spider-men just pointing at eachother, going right, and we're
of.
We're like almost like theblameless society, or we're
gonna.
We're of, we're like almostlike the blameless society, or
(23:38):
we're going to, we're notblaming, like, just tell me what
the problem is, right, and thenlet's break it down, right.
And so, as network engineers, Ithink that's part of our job is
to be that adult in the room andto be able to say okay, you're
on the application side, whatare you saying?
Show me your screen, show methe code that you're running.
And it's like you're lookingthrough their code.
(24:00):
And you go oh yeah, that youdidn't in your request, you
didn't request that URL to beallowed in the proxy, or you
know, or in URL filtering, youknow like.
And they go oh yeah, sorryabout that.
And you're like OK, hold on twominutes later, and then
everything's working Right.
Like, okay, hold on two minuteslater, and then everything's
working right Because you lookedat someone else's code, because
(24:21):
you understand Python, youunderstand Ansible, you
understand all these languages,and so being able to look at
what other folks are doing andfind that needle in the haystack
is almost invaluable.
So, you know, being that jack ofall trades is really important
and I think that's the biggestskill, like if you're
interviewing and you're, youknow, giving advice, like that's
(24:43):
one of the big pieces of advicethat I would get.
Be that jack of all trades,know a little something about
everything you know, learn howto be the adult in the room.
Speaker 9 (24:51):
Just kind of piggyback off what he was saying
.
Even getting once you're, youknow, get the DevOps into that
cloud environment, settingpolicies and procedures of
spinning up specific VMs onweekends and letting them run in
a test environment, but who'sgetting the bill?
The infrastructure team, andit's a complete disconnect.
You're spot on about thatPolicies and procedures and
(25:11):
bringing them together alongwith security in those
conversations, but that's veryvalid.
And even having to rely onceyou get to the cloud, then the
vendor teaching you AWS or Azure, or having to get up the ramp
but so many companies arerunning so lean right now they
don't have time to learn it, soit's very valid.
Speaker 1 (25:30):
When you said you don't have time to learn it.
That was my reason slash excuse.
For a really long time I wasworking three to four
maintenance windows a week.
You know what I mean?
I had little kids at home.
I was exhausted just keepingthe lights on.
They're like and, by the way,you're going to learn all this
automation.
I'm like when, right, like whenam I supposed to do that?
Now I'll take responsibility,because my manager did tell me
Tuesdays you can work onautomation learning, and there
(25:53):
was so much work and I didn'tsay no, like well, I gotta work
on scripts all day, tuesday formy Tuesday night maintenance
window.
So I take responsibility that Ishould have said no to work
that they said was supposed tobe for automation, but that I
just didn't have the time.
So it always comes down to likenights and weekends and, like
you know, oh, the kids are inbed.
Now Let me go like learn somenew stuff.
(26:14):
But got to figure out when.
Speaker 10 (26:15):
Yeah, so mentioning things where I've seen things
change over my career and beingin just enterprises for the last
20 years and from when I wascoming up in the beginning I had
to become the CLI jockey, getin the routers doing everything
that way and then eventually gotinto automation.
But that was, to Ned's point,that core understanding that I
(26:39):
built up over those 20 years ofbeing in the networks firewalls,
low balancers it helped me beable to be more proficient and
get ahead faster because all Ihad to do was learn a little bit
of Python and that propelled methat much further.
But what I noticed was some ofthe when we bring in interns or
new people trying to break intonetworking, or even somebody who
(26:59):
said they had been innetworking for a few years, they
would come in with a whole lotof ClickOps knowledge.
They only had access to a GUIor a Meraki dashboard and they
came in saying I'm a seasonednetwork engineer, but they
didn't understand what they weredoing.
(27:20):
They didn't have those 20 yearsbehind them.
So they're like oh, a VLAN, youwant a VLAN?
I know how to put that in.
I go click, click, click.
Here's a VLAN.
Okay, here's a trunk.
I turn on this port, but theydidn't understand the impacts of
what they were doing.
I saw a lot of that kind ofgoing away as that age of GUI
came along and hopefully thatstuff is starting to disappear
(27:47):
as the age of APIs are coming inmore and that stuff becomes.
That back-end knowledge becomesmore important and I think
that's one of the reasons whythere's not as many strong
engineers that were coming up inour career path right now
because of that disconnect for afew years where they had a lot
of click ops engineers that arenow being replaced with guys
coming in with that Pythonknowledge.
Speaker 4 (28:09):
So I think it's more of a question, but I guess do we
see a future where right nowthere's not a lot of young
people coming out of collegesaying they want to go into
network engineering and so whenyou're hiring a junior engineer
you know it's probably limited?
Do we see a future where we'rehiring programmers that then are
coming in and learningnetworking instead?
Speaker 3 (28:28):
I don't think so.
It's still TCPIP.
I mean, how many outages arecaused by BGP and DNS?
The TCPIP like learning how youknow all this stuff works.
It is critical.
You can't come in and spend allyour time with code and then
just haphazardly pick up howokay, like you know, even doing
(28:51):
a packet walk okay, okay, we'rebits, Now we're going to frames
and we're going in thisinterface and now we're packets.
We're going through thisrouting protocol and now we're
advertised to this.
Like that's one thing that Iused to do in job interviews.
I would actually set up a laband when people would come in, I
would have them do actuallyvery simple things, just very
(29:11):
simple things, and then youcould say, okay, on their resume
, they're an expert at this, butthen they couldn't follow
something on a single device.
They're an expert at this, butthen they couldn't follow
something on a single device.
So that tells you a lot.
So having the and it doesn't,it's not easy to learn those
things.
I have like these gigantic TCPIP books on my shelf at home
that I referenced back to a lot,Cause that knowledge leaves you
pretty quick too.
(29:31):
You really have to keep on it,you know.
So I think the counterbalanceto that is partnering with a
software developer, and this issomething I did at one of my
previous organizations that Iworked for is we had some folks
that were like high qualitysoftware developers that sat
with us for probably like 20 oftheir 40 hours and not only did
that help me I mean I thought Iwas a good coder at the time and
(29:54):
that showed me that I wasn't agood coder and there was a lot
of stuff I could improve but ithelped the team, it helped with
the knowledge sharing and it waslike a very, a very cohesive
relationship, because nobody canbe an ex.
Well, you do have folks outthere that you know, seemingly
are photographic memories andexperts at so many things, but
for the you know most of us, you, you can't be an expert at all
(30:16):
these things and you have topick what you put your time into
.
Speaker 6 (30:19):
I will say I think that in some cases maybe, yes
you will have to hire aprogrammer to do networking jobs
if not enough people are going.
What is the track now forsomeone who wants to be a
network engineer other thangoing through the Cisco training
program?
How would you decide that thatperson's a network engineer that
I want to hire?
I think if we're seeing adecline in folks coming into
(30:41):
network engineering, it meansthat the industry has to work on
maybe something more like anapprenticeship model where we
don't expect to hire somebody,you know, day one who's going to
get into the routers and dowhat needs to be done right away
and have that fundamentalknowledge We'll have to go
through.
Maybe they did a CS degree andthey're comfortable with
something like Python orTerraform, but we're going to
have to help them then get thosefundamentals and networking if
(31:03):
the discipline is going tosurvive.
Speaker 5 (31:05):
I was just talking to .
Well, just, this was like twomonths ago, but we'll say just I
was just talking to someone whorecently graduated from Temple
University with a CS degree andshe got a job at Comcast and I
think that episode justpublished last week so you can
go listen to it.
But to summarize her experience, she didn't learn any of this
(31:25):
in her CS degree.
She learned how to write in C,she learned how to like cobble
stuff together in variousprogramming languages.
She didn't learn anything aboutCI, cd or pipelines until her
final senior capstone projectand even then it was kind of
optional, like they didn't learnanything about CI, cd or
pipelines until her final seniorcapstone project, and even then
it was kind of optional, likethey didn't learn about GitHub
and I'm like come on.
So I think it's unlikely toexpect someone to come straight
(31:48):
out of college and be ready tobe a network engineer or any
other technology job aside fromthrow them into like a junior
developer role.
She was very lucky to land witha team that has a strong
culture of taking junior peopleand giving them the room to make
mistakes, helping them to learn, helping them navigate the
(32:11):
system, and she's now been ather job for a year and she's
learned a ridiculous amount ofstuff about different topics by
being curious and having aculture where she's supported to
learn.
So, as much as we want likeperfect engineers to kind of
walk in the door, it'sunrealistic and I know for those
of you who are not inmanagement or hiring you don't
(32:31):
get to decide the budgets, andif they give you enough budget
to hire one person and you'vegot all this work to do, you
want that engineer to walk inthe door and be like I hire you,
you're good, let's go.
I would hope that you would becomfortable to push back a
little bit and say I wouldprefer to hire a junior person.
It's going to be a littleharder for a while.
(32:53):
We might not be as productive,but we only have to pay them
like half and then they're goingto get really good and then you
know we'll we'll pay them moreand then we'll bring in a new
junior person that we can trainup as well.
So I think it's as much onthose of us who had been in the
industry for a while to helpcreate that pipeline and find
(33:14):
the curious people, cause I knowthey're out there.
I know there's people that arehungry to get jobs and learn a
skill.
We, as the more senior people,need to be willing to help them
along that pipeline as much aswe want to help ourselves.
Speaker 11 (33:29):
So the things that I've been stringing together
here, like there are differentemployers Expecting a fresh new
college grad to know it all andbe ready for everything that's
never been realistic, right.
And like there are legendaryprograms like Sprint had a
program called the Associate SCEprogram where you got hired and
(33:52):
then you spent 12 weeks taking12 week one week long classes,
testing and you got to pick yourassignment based on how well
you did in the program.
Cisco's had a great associatese program.
Juniper has gone up and downwith it, um, and I participated
(34:12):
in that right, um, and those areall good things, but none of
them are enough and I want toget this on the record here.
I think USNUA could make this anemphasis.
How do we do outreach to newand career people and
non-traditional?
I think that's another channelthat we've just not developed.
(34:34):
We all know so many people whogot into this without a college
education, right.
And the things that you don'tneed a bachelor's to do help
desk and to do so many otherthings right.
I think we can do more thingstogether.
I think if we could find a waywith US and UA and confederate
(34:56):
with NAF and NANAG and maybe putsome other industry programs in
place.
I don't mean this as a rebuke,but I'm tired of identifying the
problem.
I would like to start doingsomething about it, so I'll stop
there.
Speaker 1 (35:14):
I just wanted to wrap that part of it up.
My concern is that thenetworking vendors are going to
monetize the lack of peoplecoming in.
Their automation is going toget so good with their little AI
assistance.
And I don't I shouldn't havesaid little because, no, I'll be
honest with you.
I was just doing some stuff inSR Linux and they had this.
I was talking in the CLI innatural language, and it gave me
(35:35):
exactly what I needed.
That was a very differentexperience for me as a 15-year
network person.
So I see the advancement and Ithink it's much more likely for
a vendor to figure it out andhave one engineer do the work of
10 because their software isgetting so good, as opposed to
how we're going to get morepeople in.
We're going to go on a nationaltour and talk to high schools.
I think it's going to be hard.
(35:55):
Like people don't want to beplumbers, I think trade schools
are down too, so we're theplumbers of the internet, right?
So my crystal ball is vendorsare just going to monetize it
and we're actually not going tohave to plug that gap because
you're going to do the work of10 in a couple of years, because
the software is going to getthat good?
I don't know.
Speaker 3 (36:14):
Well, this is already happening in the startup space,
though, like startups areexpected with these AI agents
Okay, you have basically theknowledge of a few PhDs.
So it's shrinking the size andthe staffing of some of the
newer startups that are comingto market, and they're expected
to do way more with less,because you know that means
everybody gets more, you know,more capital at the end of the
(36:37):
run because there's not as manypeople involved.
Speaker 12 (36:39):
So I think it's all very good points about
educations and different ways tocome into the networking career
field.
When I first got into thenetworking career field, the
idea of having a Bachelor ofScience, even in being a network
engineer or even a seniornetwork engineer, was a foreign
concept for a lot of individuals.
I'd like to see us move back inthat path.
A lot of the network engineersthat I know came up from that
(37:01):
path.
They were found playing lasertag.
They were working in a lasertag store or they were working
at a gas station and they wantedto improve their life.
It was a career path to them,for you know a better way.
The thing with the AI commentand how AI is going to generate
all the commands that you need Ithink the one piece that I've
learned along the way is therestill is going to have to be
(37:24):
somebody who understands thatunderlying protocol.
I remember implementing my firstVersa SD-WAN solution when it
was on 15.2 and they had noflows or anything else to go
about it and I sat down withtheir engineers to figure it out
.
We were doing all those buzzterms to make the product better
week after week, no flows oranything else to go about it and
I sat down with their engineersto figure it out.
We're CI CD.
We were doing all those buzzterms to make the product better
.
You know week after week, butyou still had to be able to go
(37:46):
in there and understand what wasgoing on at the packet level
and how it was going out.
It was being deconstructed,being put back together.
That is where the networkengineer is going to excel in
the future.
If they're not stillunderstanding the history and
what's actually going on at theframe packet level and even
lower, we might not see as manynetwork engineers, but the level
(38:10):
of expectation of knowledgeyou're going to need is going to
be much higher, in my opinion,to be an efficient network
engineer.
Maybe.
Speaker 1 (38:16):
For the same salary?
Speaker 12 (38:17):
Yeah, for the same salary yeah, for the same salary
, it's not going to change muchyou're going to do four jobs for
the same, right.
Speaker 1 (38:21):
I mean, that was part of my resistance when all this
stuff came like I got I got tobe a network engineer, I got to
be an automation engineer, I gotto be a cloud engineer.
I got like really one salaryfor all.
But that's just, I think.
Speaker 12 (38:32):
But that's what's happening, right, that's the job
we saw that in otherengineering disciplines as well.
Think about the 1950s.
You saw drafters.
It took an entire room, or tworooms, of drafters to come up
with an aircraft.
Now you have what?
Three dudes sit down in frontof a computer and bam, now you
produce the next Boeing 777.
Speaker 8 (38:50):
This kind of dovetailing off what he said.
I mean, I think the largerquestion is how do you become a
professional?
How do you become aprofessional at anything?
Forget networks, right?
I mean, if you want to be arocket engineer, you don't go
study rockets, I mean, you studyphysics.
And how do you study physics?
You study math, and math is thebase principle of putting a
rocket.
If you want to go into themilitary, if you want to figure
(39:12):
out how war works, you go studywars from hundreds of years ago.
Right, I mean, there are baseprinciples of logistics and
things that go into codifyingmilitary science, right.
So, like, if you want to be aprofessional at anything, you
have to know what the baseprinciples of, and ultimately
the base principles of networksare electrical engineering,
algorithms, distributed systems,computer science, everything
(39:34):
else on top of it, the cli andhow you automate it.
I I mean, that's all likerecent stuff.
But if you want to, like he wassaying, like, fundamentally
understand or be a professionalat this discipline, go study
math, you know, and you'll beable to apply it over and over
and over.
I'm a, you know.
I was telling at the break,like you know, I was kind of
(39:55):
foolish enough to get two CCIEsback in the day.
But honestly, like, theundergrad in computer science is
worth so much more and amaster's degree is worth so much
more.
It teaches you how to think andif you know how to think, you
can apply it over and over andover, like just knowing.
Cisco CLI is a fad.
Speaker 5 (40:13):
I think about how young the tech industry is as a
whole.
I mean, you know, plumbers havebeen around for hundreds of
years.
That's a trade that's beenaround for a little while.
Think about all the otherindustries out there that are
hundreds or even older.
Tech is what?
Like 70 years old if we want tobe generous, 40 years old if we
want to be slightly lessgenerous.
(40:33):
So if that's how long we'vebeen doing it, there's still a
lot more to learn.
We're in the infancy oftechnology and I think you're
right to a certain degree.
Just having fundamentals thatare beyond technology and going
into concepts that never get old, like math basically, and being
(40:55):
well-versed in that, willguarantee that you can find a
spot somewhere.
But I don't think everybodywants to do that.
I think you're going to havepeople who are comfortable being
generalists and for thelisteners I pointed at myself.
I came from help desk, I wentinto sysadmin, but I worked at a
(41:16):
small company, so we wereresponsible for everything.
When I said everything, if theair conditioner broke, I had to
call the repair guy to come outand fix the air conditioner.
If the vending machine wasempty, I had the keys to go open
the vending machine and so,coming from that, I had to
understand how a storage arrayworked, how networking worked,
how VMware worked and also someweird stuff about Microsoft Word
(41:40):
that I don't even want to getinto.
But I had to have that sort ofgeneral knowledge and I was
comfortable with that.
I've never been a go deep onone thing and just be an
absolute expert in that onething and that's just.
I know that's me, that's theway I am, and I think that's
kind of where things are headed.
You can either be a generalistin the cloud or a generalist
with technologies, or you can gosuper deep on one thing.
(42:03):
Vendors are going to eat outthe middle, they're going to go.
We can make it good enough.
We can make a Meraki where it'sclose enough and easy enough to
click ops your way through it,that any small business or even
medium business never needs toopen a CLI because they can just
click through and connect allthe cables and it just works.
And if you're a generalist,that's perfect.
(42:23):
And when something goes wrongthey can call the network
engineer at Meraki, who is thedeep expert, and they'll figure
out what's actually going on.
And I think that's the caseacross all technologies, not
just networking is that softmiddle where we needed someone
with those skills.
That might be shrinking and youeither have to go deep or wide.
Speaker 13 (42:43):
I've been working computers for a very long time.
When I got into it there was noGoogle, so it's funny seeing it
grow.
So it's like CLI, that's allthere was back then.
I've grown up through computers.
My son's now 25, working for apretty big VAR and he's kind of
learned similar stuff Also onthe network routes.
He's doing segmentation andit's funny seeing the
(43:04):
differences.
So where I grew up with CLI hewas the opposite.
He's kind of gone the gooeyroute because that's what
they've learned.
And as someone also who's hiredpeople out of college, what I'm
kind of seeing is the tools aregetting better for networking
and you're going to see a lot ofways to make what you used to
do back in the early 2000s andeven before that.
(43:25):
It takes a very long time.
Things can get quicker, butwhen it really breaks those
tools are not going to be asgood and that's where you kind
of need the people who do knowwhat they know.
The people that worked at thebottom of the OSI stack and up
the way up know that stuff a lotbetter than the people who
start at the top and work theirway down.
Speaker 6 (43:43):
Yeah, but that's something that comes with time
and experience.
Speaker 13 (43:45):
Yes, so yeah, and that's the thing.
So it's like when, when I hired, we hired a kid out of college
great worker, he is a juniorperson now sent him to send him
to training so he getsexperience, which is great, he
learns his stuff, but the bigthing is he's shadowing the
people that they they may havelearned the books, but they also
learned where it's important inthe real world, because the
(44:07):
books don't always tell thetruth in a way, but that that's
yeah.
That's where they have to learnis is the younger kids need to
work with the older ones thathave been around so they can
then see where you know itdoesn't always work the same way
.
Speaker 2 (44:21):
So we we have about two minutes left, Um, and then I
think we hit an hour, which youguys are the experts, but my
understanding is you go over anhour on a podcast.
People stop listening, that'swhat.
Speaker 9 (44:30):
I'm told anyway.
Speaker 2 (44:37):
So I want to be cognizant of that, but I did
want to pause, you know, ask forclosing thoughts just relative
to the conversation.
Right, Much like our past nugs,we got through a total of one
question today, which is great.
Just love to do that.
It's good because it means thateverybody's involved.
Right, the conversation'shappening and that's really what
we want when we're here.
But just want to ask forclosing thoughts before we kind
(44:58):
of move into more of a wrap-upand we kind of close the event
out here.
Speaker 6 (45:01):
Yeah, I'll start.
First, I'm excited that therewas so much back and forth, like
we didn't just want to come up,so I appreciate everybody
weighing in with theirexperiences.
That was great.
It's great to have more of aconversation.
Second, I'm excited about thefuture of networking.
I think there's a ton ofinteresting stuff.
So, ladoo, really fascinatingproblems to solve and I think if
we pitch it the right way, wecan bring in a new generation of
(45:22):
network engineers and catchtheir curiosity and give them
all these cool tools and letthem go play and do incredible
things.
So I'm rosy about the future ofthis discipline.
Speaker 1 (45:32):
It's a great career, right, I mean it's a great job.
Nobody likes change.
I saw a lot of change.
I didn't like it.
I cried, Nobody cared, Got laidoff, panicked, learned a bunch
of new stuff and now I'm in abetter spot.
You know what's different fromnow than my 2012 CCNA?
Learn Linux, learn automation,learn cloud.
I think if you can learn thefundamentals of networking and
(45:56):
those other things, I thinkyou're in really good shape.
Speaker 3 (45:58):
I feel compelled to make my closing thoughts about
AI.
So seriously, though I almostcame like with the AI stuff I
was what I'm seeing it's kind ofscary almost is one of the big
(46:20):
skills over the years that Ilearned from not having Googled
everything and having an answerimmediately available.
Was it kind of the gentleman'spoint back there?
It helped me think and workthrough problems like I had to
figure it out.
I couldn't just type into chatgbt how do I do this?
And get an answer immediately.
Because when you do that foreverything, you find that's how
(46:41):
you train your brain.
That is the training thatyou're taking in.
You're not actually building askill set and that's something
that I honestly kind of scaresme.
But I know that AI is going tobe a big part of the future and
I was just in New York and Idon't know if you all know who
John Capobianco is.
He is really bullish on AI andhe actually made me excited
because he sees the good andwhere it's going and he kind of
(47:02):
narrows into where the realvalue is and is super excited
about it.
So that's just a reminder to methat all these new technologies
about it.
So that's just a reminder to methat all these new technologies
, you can't discount any of them.
You really have to see wherethey're going to be practical.
But again, how you use that chatGPT, like I know that I could
go in and say, hey, I want youto write this code for me, but I
(47:24):
still make my it's hard, but Istill make myself.
Not do that.
I might give it some of my codeand say, hey, what is wrong
with this?
But don't give me the answer.
I don't want you to give meanything back and I can get some
feedback and then go and figureit out with less stuff.
And I know that's still kind ofcheating, but at the same time
I still want to keep sharp so Ican actually write things and I
(47:44):
know how it works and I know howeverything you know connects
together.
So yeah, just I guess myfeedback and closing argument is
you know there's a lot of stuffcoming our way and you only
have so much time.
You know, try to prioritize anddo what's best for you and your
role.
You know what you're doing withyour job, cause if you can
apply it practically, you knowit's going to stick better and
(48:07):
it's going to serve you betterlong-term.
Speaker 2 (48:09):
With that we'll.
We'll close it out there.
Again, appreciate everybody'stime.
Thank you for the panelists.
Hey folks, back in October 2024, I had the pleasure
of participating in USNUA'sfirst podcast panel, where Drew
Conroy-Murray, ned Belovance,william Collins and I discussed
the current state of the networkengineering profession and what
it takes to create and sustaina successful career in our
ever-evolving field.
The fine folks at USNUA haveallowed the art of network
(00:24):
engineering to share the showwith you on our platform and I
think you'll enjoy the dynamicconversation that follows, thank
you.
They host networking usergroups, or NUGS, all over the
country, where free food, freebeer and free information is
(00:46):
presented in a relaxedatmosphere.
For more information on theUSNUA, as well as access to a
variety of member-onlyeducational resources, head over
to usnuacom and check out themembership section to become a
member of this fantastic groupof network engineering
professionals today.
Now sit back.
Enjoy our PA Nugs podcast panelfrom October 2024.
(01:09):
As always, thanks for listening.
Speaker 2 (01:16):
I think I mentioned when we started this thing.
This is the first time we'redoing something like this.
It's a joint initiative betweensome different people from the
community and the USNUA toreally bring together what we're
referring to as the podcastpanel, right?
Just basically people that areinvolved in the community,
people who are in the know andpeople who have a pulse on the
industry, and what we're tryingto do is really get their
(01:38):
perspective, you know,essentially put a face to the
voices that everybody'slistening to on these podcasts.
Right, and I'll let everybodyback here behind me give an
introduction for themselves.
So I don't butcher it from myside of things.
But the one thing I did want tomention as we get into this is
if you ask a question, I'mprobably going to repeat it.
They mic'd me up for somereason so that way they can
record everything.
I don't know podcasters, weirdpeople.
(02:00):
I also have a second mic.
If you have a comment, chancesare I'm going to pass you a mic.
So if there's a comment, feelfree to, like you know, wave
your hand, throw something at me, whatever's easiest, and if
it's a question, I'll probablyjust repeat it so that way they
can respond to it.
I passed around a couple ofquestion sheets.
I guess we'll call them.
If there's something on thereyou really want to talk about,
you know, feel free toacknowledge it.
(02:21):
Right, you know, we'll getthere in some way, shape or form
.
I'm probably going to startwith changing landscape and
we'll go from there, but we'lldive in.
But, eric, if you want to go tothe next one for me, there we
go.
I'm going to go left to right,although the pictures don't line
up that way, and I'm going tolet everybody introduce
themselves here real quick andwe'll dive in big enterprise.
Speaker 3 (02:52):
Started out, my first job was actually managing CSS,
load balancers, certificates,which was like the worst job
ever.
I would have rather worked fastfood at the time, but you know
I put in the time.
So I started in the data centerand then pivoted to cloud.
When cloud started reallybecoming a thing, did the
multi-cloud thing, solved a lotof problems in the enterprise.
And then now I work for astartup for three years my first
(03:14):
time working for a vendor andthen I started the cloud gambit
podcast and I co-host withYvonne Sharp and really what we
do is we focus on a lot of theadjacent areas to network
engineering.
So my core was networkengineering.
But one of the cheat codes I hadwhen I was in enterprise I'm
curious, I like talking topeople.
(03:36):
So that was how I was able toget things over the finish line
in the enterprise.
Hey, I'm gonna go to talk tosecurity, I'm gonna talk to the
developers, and it's some ofthose points like we couldn't
work from home.
This is a while back.
And so, hey, they're in theoffice.
I just go over and talk to them.
You know they're right there.
So I figured, hey, why not dothat with some of these engaging
conversations.
(03:56):
So we talked to everybody from,like network engineers,
founders of startups, venturecapitalists, because it all,
like, really is closely meshedtogether, so happy to be here.
Speaker 1 (04:07):
Hey guys, andy Laptev , co-founder, co-host, art of
Network Engineering.
I started out in tech workingat a Verizon Central office
pulling wires for POTS lines andDSLs and getting the bejesus
shocked out of me when you wouldtouch the actual things.
Cable guy Comcast for fiveyears worked my way into their
NOC with a CCNA and I got a jobin fintech for six years
(04:30):
managing their global WANs.
Burnout there and right towardthe end of my burnout got
plucked at Juniper to do someproduct management stuff, some
customer experience work andwhat else can I tell you?
Our show, I guess, because thisis a podcast panel, we wanted
to tell the stories of networkengineers doing the job.
So people like you, theoperators we didn't really think
(04:53):
that audience was all that wellserved at the time.
So you know what's it likeworking in an Ock, what's it
like climbing that ladder,what's it like being a network
engineer, and it's been awesome.
I've gotten to meet a ton ofpeople.
I mean we were just at CiscoLive.
I went to my first Cisco Live.
I couldn't walk 20 feet withoutsomebody being like oh my God,
are you the guy from the thing?
I'm like whoa.
So it's just been beyond mywildest dreams.
(05:13):
Started in COVID.
It wasn't like a plan, we werejust home, we were studying for
start a podcast.
These conversations in ourstudy group are kind of funny
and it just kind of took off.
But great to be here.
Thanks everybody for coming.
Thanks Andy.
Speaker 5 (05:28):
Hi, I'm Ned Belovance and let's see, we're doing
history here.
So it all started on a warmsummer day in 1970.
No, so I got my start in techon Helpdesk as I think a lot of
people might and worked my wayup through there to be a desktop
(05:49):
engineer.
And then I started getting intonetworking, took my CCNA, tried
to go for my CCNP, failed thefirst time, said forget this
stuff, I'm going to learn VMware, and never looked back.
So became a systemsadministrator and then a
consultant, and I worked inconsulting for eight years and
while I was doing that, cloudhappened.
(06:09):
So I had to start learning AWSand Azure, because that's what
our clients wanted us to setstuff up on.
And hey, networking is there.
So now I have to remember allthe stuff I learned in CCNA and
also then augment that withadditional things that were
unique or somewhat unique to thecloud.
So that's what I was doing foreight years, and then I decided
(06:32):
to get out of the consulting bizbecause that is a lot of stress
Stress, I think, is the rightword I'm looking for and so I
went into full-time contentcreation and being a technical
educator.
So I have a bunch of courses onPluralsight around things like
Terraform, vault, azure, and Ialso teach live stuff.
And, oh, I have a podcastbecause this is the podcast
(06:54):
panel.
I have two podcasts because oneis never enough.
Oh God, they're like Fritos.
So, seriously, I'm on like mysixth podcast at this point.
So one is called Day Two DevOps.
That is on the Packet PushersNetwork that I host with Kyler
Middleton.
She's awesome.
We talk about anything looselyrelated to DevOps.
(07:15):
So that could be careeradvancement, it could be a new
tool, could be just someonewho's got a cool thing they're
doing at work and they want totalk about it.
And I also host a podcast calledChaos Lever with my buddy,
chris, where we talk about thehistory of technology.
So we recently did a podcast.
We recently covered that wholeWordPress thing that's happening
(07:36):
.
I don't know how many peopleare aware of the nonsense that's
happening with WordPress.
I don't really care about that,but it gave me a chance to talk
about BBSs and blogging and howall of that morphed into the
fact that WordPress runs like50% of websites now.
So if you like tech history alittle bit, should I give them a
little preview?
Go for it.
I already forgot what I'mworking on for next week.
(07:58):
That's terrible.
I spent all day writing no, I'mkidding, excel is turning 40.
Microsoft Excel is turning 40.
Microsoft Excel is turning 40.
And it also runs like 70% offinance in the world.
So I thought it'd be fun tolook back at spreadsheets a
little bit.
Also, like 90% of automation,most networking is really just
(08:19):
configured through a spreadsheet.
So I thought it'd beinteresting to look back at the
past of spreadsheets where theycame from and how they evolved.
Speaker 6 (08:27):
And that's me.
I'm Drew Connery Murray.
I am kind of the odd person outhere.
I came into this industry onthe reporting side, so I've been
covering the tech industry as areporter, writer, editor,
blogger, podcaster for almost 30years now 20, 25 years now.
I've worked for outfits likeNetwork Magazine, network
(08:52):
Computing Information Week.
For a while, when the Interopconference was running, I was
the director.
I was responsible for bringingin all the technical
presentations that happened atInterop.
Through my role as an editor atone of the magazines I was
working for, I got to know EthanBanks and Greg Farrow, who
(09:14):
founded the Packet PushersNetwork.
Greg invited me to start apodcast with him called Network
Break, which is like a weekly ITnews show.
We started it.
We were having a good time.
Then I got laid off fromInterop.
So I went to Greg and was likeif you want to find another
co-host, you know somebodyconnected in the industry,
that's fine.
He's like do you want a job?
I was like, well, yeah, I needa job.
That's how I joined PacketPushers.
(09:34):
I've been there many years nowand I co-host three podcasts
Network Break.
I co-host Heavy Networking,which is sort of the flagship
networking podcast, and a recentpodcast that I started with
Jennifer Minella called PacketProtector, which sits at the
intersection of networking andsecurity.
Ned is on the Packet PushersNetwork and Scott Rabban has
Total Network Operations.
(09:55):
We just launched a new podcastcalled Technically Leadership.
If you're looking to get intotechnical leadership, that's a
great show and we've got somereal niche ones like IPv6 Buzz
If you're trying to get yourhands around IPv6, some really
deep content there.
Speaker 2 (10:09):
Given your perspectives right and the
history that all of youmentioned here during the
introductions, the landscape innetwork engineering today right.
Maybe we just start therePerspective of where you've seen
it come from, where it's goingand what you think the next
generations are going to looklike.
For today's network engineers,what they need to know, where
the industry's heading.
Speaker 3 (10:29):
Yes, I remember when I first started doing any sort
of automation it was a long timeago Bash, perl, expect, back to
the hating CSS load balancersthat was what I was using to
solve that problem.
I wanted to automate it and itfreed up a lot of time.
It allowed me to work on otherthings and back then it was a
(10:49):
combination of some differentscripts and we basically had a
jump server and in that jumpserver we gave the developers
the ability to fill in a certainspreadsheet.
They would put it in adirectory and then I had a cron
job run at strategic timesduring pre-approved change
windows that would go out and dothings.
When you look at that and youlook at, okay, it's Linux, it's
(11:10):
a cron job, you're gluing thingstogether.
And then I look at recently,these cloud pipelines I've built
.
It's so eerily similar becauseit's okay, I'm still using Linux
, I'm still gluing thingstogether, I still have this,
this instead of a long-livedserver.
It's ephemeral, like it'scoming up during the pipeline
and it's getting everythingstaged and then it's, you know,
(11:31):
tearing back down.
But there's so manysimilarities there and just
thinking about how much time haspassed and looking at okay,
that was then.
Okay, it was pearl and bash,but now, okay, it's still bash,
and maybe it's Terraform andPython, or Terraform and Golang,
and you know a YAML file thatsits in a Git repo.
It's very similar.
So, yeah, things have changed.
(11:53):
And I think the thing thatreally has changed is you have
networking under the hoodeverywhere.
So even if you look at you knowyou have physical, you go up to
layer three and layer four andthen you go up to the
application networking.
But even as you layer on moreabstractions, under the hood
(12:15):
it's a lot of tunnels, a lot ofreverse proxies and a lot of
things.
But the thing that's actuallychanged and one thing that I
find a lot of network engineersare, I don't want to say, scared
of, but they just want to stayaway from it is the tools that
we use to interact with thesethings.
So, instead of going in andhammering on various CLIs, the
workflows change.
The way that we interact withthese things has changed and
architecture's changed too.
(12:35):
So with the more abstractionlike the other day, I helped
someone that.
So it was a Kubernetes clusterin AWS add an EKS cluster and
they made a change to someTerraform.
I think they changed theversion or something and that
actually tore the clustercompletely down and brought it
back up.
They caused an outage, like theimmutable aspect of this.
(12:56):
But there's a lot of things youcan do that will actually make
configuration changes to thatcluster in place, like maybe a
node group or tags or thingslike that, where you're not
gonna cause an outage.
So really in the architectureworld in the cloud, it's a lot
of reading, it's it is kind ofdry, but you just have to know
how these resources work andwhat their behavior is you
(13:17):
embraced automation way beforeme, or anybody else I know did
so.
Speaker 1 (13:22):
I got my CCNANA in 2012, I think, and there was no
automation on the exam.
There was none being taught.
I got a job at a NOC there wasno automation at all, and five
of six and a half years at theFinTech role there was no
automation.
To speak of what's changed,like to answer your question
what's changed?
Automation and cloud were thetwo big 800 pound gorillas that
(13:43):
kept peeking around the cornerand I kept going get the hell
away from me, I want nothing todo with you.
I am the CLI jockey, right,like that's the job.
And I was brought up maybe in adifferent time.
That was the whole program.
I went to Netacad Like that wasthe job.
So fast forward, years later Ifind myself.
I fought all the changinglandscape as long as I could and
then I thought if I could runaway to a vendor and just smile
(14:08):
and be charming and helpcustomers be happier, like, oh,
this is great, I don't have tostudy anymore and screw that
automation stuff.
And then I got laid off inDecember and I started looking
at job postings.
I'm like shit, they all wantautomation.
You have to know Python orAnsible Infrast.
All want automation.
You have to know Python orAnsible Infrastructure is code.
You have to be versed in acloud.
(14:28):
So really, for me it was onlyout of.
I was beaten in a submission iswhat I'll tell you.
The industry has changed.
If you look at job postingstoday as opposed to even five
years ago, I have a lot ofpeople come to me Because we do
a lot of work in the communityhelping people, and I probably
have seven people right now I'mtrying to help with resumes or
trying to get into the industry.
What are they doing?
And what I told people five orsix years ago get your CCNA, do
(14:52):
some content creation so you canshow your soft skills, your
communication skills, and builda home lab.
It doesn't have to be physical,but you're not going to have
experience and people are goingto call you out in interviews
and a home lab can show you haveexperience.
We don't have any productionexperience.
Well, the hell, I don't.
I got four 2911s, I got 3750sWhatever right and I got a
server.
So it's a way that I got aroundthe no experience thing.
(15:15):
So I say all that to say thatthe past I don't know.
Six or seven months I've beenlearning Python.
I've been learning GitHub.
I haven't gotten into cloud yet, but that's kind of next on my
radar.
I'm reading an LLM book rightnow that Phil Gervasi just
turned me on to, so I'm tryingto get caught up in all the
things I avoided all those years.
There's nothing likeunemployment to wake you up out
(15:37):
of your cognitive bias andre-change your compass of like,
oh, just because I don't want tolearn automation.
Nobody gives a damn.
You know what I mean.
And if you're at a job that youdon't have to do any of that,
good for you.
I was in a job.
I didn't have to do any of that.
I'll end with this.
The job that I had for sixyears that I was a senior at
Edge ran all the WAN projects.
They wouldn't hire me Pythonknowledge.
(16:06):
So I can't even get the jobtoday that I got before at the
same company.
I was a rock star there.
Just because I'm not Pythonproficient, I'm getting there.
I just ran my first script inNetMeco and I logged into a
router and did a show for inwhatever my just visual studio
code.
And I'm also learning Linux.
Right, these things that, likeit was on Scott's thing earlier
Linux automation, you know allthat stuff.
So it is changing and I'm justbusting my ass now trying to
(16:28):
catch up.
Speaker 5 (16:30):
I'm the cloud guy, so I've got to say that the
biggest change has been therollout of the cloud.
And it's not just thathyperscalers exist, but they put
a pressure on the way thateverything else is done Because
they're the 800-pound gorilla,they're the 2,000-pound gorillas
in the room.
Else is done because they'rethe 800 pound grill.
They're the 2,000 poundgorillas in the room.
They're the ones who can drivethe design of specific ASICs
that are going into network gear.
(16:50):
They can determine how systemboards are designed for the
compute that they're running.
They have a lot of influence, Iguess is what I'm saying.
And their philosophy iseverything needs to be
programmable, everything needsto have an API and everything
needs to be automation friendly,because there's no way we can
do this at scale If somebody hasto click on next, next, finish.
(17:11):
And you know, I came fromdesktop support and a world of
Microsoft where a lot of stuffwas next, next, finish, and I
also had to learn automationslightly different context, but
same basic thing.
I was managing Exchange 2003back in, you know, 2003.
(17:32):
And then Exchange 2007 came outand, for those who don't know,
exchange 2007 was theintroduction of PowerShell into
the Microsoft ecosystem for real, and there were certain things
you could only do in Exchange2007 through PowerShell.
You couldn't go into the GUIanymore, and that was great for
(17:54):
big organizations who were likewe want to run these scripts at
scale and do things, and itsucked.
When you're an SMB and suddenlyyou are also responsible for
writing PowerShell, you canguess where I was.
But it forced me into the worldof automation and by doing that
I learned PowerShell, which isscripting, programming, call it
whatever we want.
(18:15):
But then I was able to applythat skill to VMware and then I
was able to apply it to Azure.
When that rolled out and then,after realizing that using
PowerShell to control Azure iskind of a pain, I discovered
Terraform and I was like, ohokay, so I can do this
declaratively now.
But the core premise ofTerraform and all these other
tools is the fact that there'sgoing to be an API you can talk
(18:37):
to.
So I guess the biggest changeis the fact that you have to
shift the mentality a little bitaway from the CLI and on to how
do I interact with an API orwrite a tool that knows how to
talk to the API?
And that's been a struggle forme as much as it's been for any
network engineer, but slowlywe're all getting there right?
Speaker 6 (18:57):
Yeah, it's hard to disagree with anything that's
been said.
I think, as Scott mentionedearlier in his presentation,
networking sort of fundamentalshave kind of we're not fighting
about TCPIP versus ATM orwhatever.
We're not battling over tokenring versus Ethernet.
That's been decided already andnow we've got that structure
and we're building on top of itin a data center.
(19:17):
It's LeafSpy and EVP and VXLANand we don't need to argue about
it anymore.
Where things are going is towardautomation, is toward more
software, developer-likeconstructs like GitHub, like
APIs, like being comfortablewith Python, ansible and these
new tooling things.
Cli skills that you put a lotof work into may not be as
relevant, but the thing that isstill relevant is all of that
(19:42):
fundamental information that youhave about how IP works, about
how Ethernet works, aboutsubnetting, about addressing.
That's still essential,regardless of whether you're
doing it in the cloud, on-prem,through some new software
interface or right into the CLI.
So it's not an abandon all hope.
You've got the fundamentalsthat are essential now, like
this gentleman from Oracle said,oracle doesn't have a cloud
(20:02):
business without a network.
Speaker 5 (20:03):
Somebody had mentioned I think it was, I
forget which one of you, and I'msorry.
Staffing is a major concernFinding people that have the
fundamental skills and that whenyou onboard someone, it's more
about a history lesson becausethey don't know that stuff.
Maybe they're an expert inPython but they have no idea how
to set up a trunk for VLANs.
(20:24):
They're like a V, what I'm notfamiliar with the terminology,
and so if you already have thosecore skills around networking,
you understand the fundamentals.
Like programming is hard butit's not that hard and I mean
I've seen people that havelearned programming.
I guarantee you can all learnit.
Um, what's really really hardis the last 10, 15, 20 years
(20:46):
you've put into learning networkfundamentals and that's
something that's irreplaceableand someone can't just
accelerate that in a three-monthcourse and suddenly understand
all the fundamentals.
So you got to leg up on people.
That way it it's just you haveto continue to learn and grow
your skillset to accommodate thetools that are driving that
underlay.
Speaker 2 (21:05):
So I do want to pass the mic around the room here a
little bit as well.
Right, thinking about wherethey've seen things come from to
where they think things aregoing.
Anybody with a strong opinionin the room?
Just relative to some of whatthey mentioned?
It sounds like cloud andautomation, two of the big ones.
Right, I'm sure we can allagree we're seeing some of that
within our roles and what we'redoing now.
(21:26):
But is there anything elseyou're seeing outside of those
two realms that you think ispertinent to what we're talking
about when we look at wherenetwork engineering or
engineering in general is headed?
Speaker 7 (21:34):
So one of the things that I've seen, I've experienced
, is we went through a huge datacenter migration right, and
part of it is networking rightand you're part of that team to
migrate apps from on-prem to nowinto AWS right, and you're kind
of forced into the cloud worldand, like you said, there's
(21:57):
networking component, transitgateways, and you're also, as a
network engineer, expected tounderstand security right, and
so that's a skill that you knowwe haven't touched on tonight,
but you have to know it right,because there's networking in
every firewall that you touchright.
There's BGP, there's OSPF right,and if you're going to do
(22:19):
direct connect with AWS or anyof these other cloud providers,
you have to understand thatstuff right and you have to
understand how you're going to.
You know peer with othernetwork, you know with your AWS
cloud and everything like that.
So I think that's one of thethings that I've experienced is
that traditional networking is agreat base when you have to
(22:41):
learn cloud skills.
Once you get into the cloud andyou're like, oh, these are
things I've done for the last 20years, it's that much easier.
Speaker 2 (22:49):
One of the interesting things about the
networking industry and I'msaying this as not one of the
panelists so cut me off.
If Andy later on edited it out,go ahead.
So cut me off.
If, like Andy later on editedout, go ahead.
But at times you almost have tobecome, because the network is
touching everything right andyou have people that are
specific to individual domains,like you mentioned.
You know you're going to havesecurity, you're going to have
(23:09):
infrastructure, you're going tohave network, so on and so forth
, but you almost have to be insome way, shape or form, kind of
like that jack of all trades.
Speaker 7 (23:15):
You do Network or form, kind of like that jack of
all trades.
You do and you are, and and forlack of better term, you're
sometimes you.
You get asked to join a call,right, and you're the adult in
the room.
You are, you're the adult inthe room because it's seven
spider-men just pointing at eachother, going right, and we're
of.
We're like almost like theblameless society, or we're
gonna.
We're of, we're like almostlike the blameless society, or
(23:38):
we're going to, we're notblaming, like, just tell me what
the problem is, right, and thenlet's break it down, right.
And so, as network engineers, Ithink that's part of our job is
to be that adult in the room andto be able to say okay, you're
on the application side, whatare you saying?
Show me your screen, show methe code that you're running.
And it's like you're lookingthrough their code.
(24:00):
And you go oh yeah, that youdidn't in your request, you
didn't request that URL to beallowed in the proxy, or you
know, or in URL filtering, youknow like.
And they go oh yeah, sorryabout that.
And you're like OK, hold on twominutes later, and then
everything's working Right.
Like, okay, hold on two minuteslater, and then everything's
working right Because you lookedat someone else's code, because
(24:21):
you understand Python, youunderstand Ansible, you
understand all these languages,and so being able to look at
what other folks are doing andfind that needle in the haystack
is almost invaluable.
So, you know, being that jack ofall trades is really important
and I think that's the biggestskill, like if you're
interviewing and you're, youknow, giving advice, like that's
(24:43):
one of the big pieces of advicethat I would get.
Be that jack of all trades,know a little something about
everything you know, learn howto be the adult in the room.
Speaker 9 (24:51):
Just kind of piggyback off what he was saying
.
Even getting once you're, youknow, get the DevOps into that
cloud environment, settingpolicies and procedures of
spinning up specific VMs onweekends and letting them run in
a test environment, but who'sgetting the bill?
The infrastructure team, andit's a complete disconnect.
You're spot on about thatPolicies and procedures and
(25:11):
bringing them together alongwith security in those
conversations, but that's veryvalid.
And even having to rely onceyou get to the cloud, then the
vendor teaching you AWS or Azure, or having to get up the ramp
but so many companies arerunning so lean right now they
don't have time to learn it, soit's very valid.
Speaker 1 (25:30):
When you said you don't have time to learn it.
That was my reason slash excuse.
For a really long time I wasworking three to four
maintenance windows a week.
You know what I mean?
I had little kids at home.
I was exhausted just keepingthe lights on.
They're like and, by the way,you're going to learn all this
automation.
I'm like when, right, like whenam I supposed to do that?
Now I'll take responsibility,because my manager did tell me
Tuesdays you can work onautomation learning, and there
(25:53):
was so much work and I didn'tsay no, like well, I gotta work
on scripts all day, tuesday formy Tuesday night maintenance
window.
So I take responsibility that Ishould have said no to work
that they said was supposed tobe for automation, but that I
just didn't have the time.
So it always comes down to likenights and weekends and, like
you know, oh, the kids are inbed.
Now Let me go like learn somenew stuff.
(26:14):
But got to figure out when.
Speaker 10 (26:15):
Yeah, so mentioning things where I've seen things
change over my career and beingin just enterprises for the last
20 years and from when I wascoming up in the beginning I had
to become the CLI jockey, getin the routers doing everything
that way and then eventually gotinto automation.
But that was, to Ned's point,that core understanding that I
(26:39):
built up over those 20 years ofbeing in the networks firewalls,
low balancers it helped me beable to be more proficient and
get ahead faster because all Ihad to do was learn a little bit
of Python and that propelled methat much further.
But what I noticed was some ofthe when we bring in interns or
new people trying to break intonetworking, or even somebody who
(26:59):
said they had been innetworking for a few years, they
would come in with a whole lotof ClickOps knowledge.
They only had access to a GUIor a Meraki dashboard and they
came in saying I'm a seasonednetwork engineer, but they
didn't understand what they weredoing.
(27:20):
They didn't have those 20 yearsbehind them.
So they're like oh, a VLAN, youwant a VLAN?
I know how to put that in.
I go click, click, click.
Here's a VLAN.
Okay, here's a trunk.
I turn on this port, but theydidn't understand the impacts of
what they were doing.
I saw a lot of that kind ofgoing away as that age of GUI
came along and hopefully thatstuff is starting to disappear
(27:47):
as the age of APIs are coming inmore and that stuff becomes.
That back-end knowledge becomesmore important and I think
that's one of the reasons whythere's not as many strong
engineers that were coming up inour career path right now
because of that disconnect for afew years where they had a lot
of click ops engineers that arenow being replaced with guys
coming in with that Pythonknowledge.
Speaker 4 (28:09):
So I think it's more of a question, but I guess do we
see a future where right nowthere's not a lot of young
people coming out of collegesaying they want to go into
network engineering and so whenyou're hiring a junior engineer
you know it's probably limited?
Do we see a future where we'rehiring programmers that then are
coming in and learningnetworking instead?
Speaker 3 (28:28):
I don't think so.
It's still TCPIP.
I mean, how many outages arecaused by BGP and DNS?
The TCPIP like learning how youknow all this stuff works.
It is critical.
You can't come in and spend allyour time with code and then
just haphazardly pick up howokay, like you know, even doing
(28:51):
a packet walk okay, okay, we'rebits, Now we're going to frames
and we're going in thisinterface and now we're packets.
We're going through thisrouting protocol and now we're
advertised to this.
Like that's one thing that Iused to do in job interviews.
I would actually set up a laband when people would come in, I
would have them do actuallyvery simple things, just very
(29:11):
simple things, and then youcould say, okay, on their resume
, they're an expert at this, butthen they couldn't follow
something on a single device.
They're an expert at this, butthen they couldn't follow
something on a single device.
So that tells you a lot.
So having the and it doesn't,it's not easy to learn those
things.
I have like these gigantic TCPIP books on my shelf at home
that I referenced back to a lot,Cause that knowledge leaves you
pretty quick too.
(29:31):
You really have to keep on it,you know.
So I think the counterbalanceto that is partnering with a
software developer, and this issomething I did at one of my
previous organizations that Iworked for is we had some folks
that were like high qualitysoftware developers that sat
with us for probably like 20 oftheir 40 hours and not only did
that help me I mean I thought Iwas a good coder at the time and
(29:54):
that showed me that I wasn't agood coder and there was a lot
of stuff I could improve but ithelped the team, it helped with
the knowledge sharing and it waslike a very, a very cohesive
relationship, because nobody canbe an ex.
Well, you do have folks outthere that you know, seemingly
are photographic memories andexperts at so many things, but
for the you know most of us, you, you can't be an expert at all
(30:16):
these things and you have topick what you put your time into
.
Speaker 6 (30:19):
I will say I think that in some cases maybe, yes
you will have to hire aprogrammer to do networking jobs
if not enough people are going.
What is the track now forsomeone who wants to be a
network engineer other thangoing through the Cisco training
program?
How would you decide that thatperson's a network engineer that
I want to hire?
I think if we're seeing adecline in folks coming into
(30:41):
network engineering, it meansthat the industry has to work on
maybe something more like anapprenticeship model where we
don't expect to hire somebody,you know, day one who's going to
get into the routers and dowhat needs to be done right away
and have that fundamentalknowledge We'll have to go
through.
Maybe they did a CS degree andthey're comfortable with
something like Python orTerraform, but we're going to
have to help them then get thosefundamentals and networking if
(31:03):
the discipline is going tosurvive.
Speaker 5 (31:05):
I was just talking to .
Well, just, this was like twomonths ago, but we'll say just I
was just talking to someone whorecently graduated from Temple
University with a CS degree andshe got a job at Comcast and I
think that episode justpublished last week so you can
go listen to it.
But to summarize her experience, she didn't learn any of this
(31:25):
in her CS degree.
She learned how to write in C,she learned how to like cobble
stuff together in variousprogramming languages.
She didn't learn anything aboutCI, cd or pipelines until her
final senior capstone projectand even then it was kind of
optional, like they didn't learnanything about CI, cd or
pipelines until her final seniorcapstone project, and even then
it was kind of optional, likethey didn't learn about GitHub
and I'm like come on.
So I think it's unlikely toexpect someone to come straight
(31:48):
out of college and be ready tobe a network engineer or any
other technology job aside fromthrow them into like a junior
developer role.
She was very lucky to land witha team that has a strong
culture of taking junior peopleand giving them the room to make
mistakes, helping them to learn, helping them navigate the
(32:11):
system, and she's now been ather job for a year and she's
learned a ridiculous amount ofstuff about different topics by
being curious and having aculture where she's supported to
learn.
So, as much as we want likeperfect engineers to kind of
walk in the door, it'sunrealistic and I know for those
of you who are not inmanagement or hiring you don't
(32:31):
get to decide the budgets, andif they give you enough budget
to hire one person and you'vegot all this work to do, you
want that engineer to walk inthe door and be like I hire you,
you're good, let's go.
I would hope that you would becomfortable to push back a
little bit and say I wouldprefer to hire a junior person.
It's going to be a littleharder for a while.
(32:53):
We might not be as productive,but we only have to pay them
like half and then they're goingto get really good and then you
know we'll we'll pay them moreand then we'll bring in a new
junior person that we can trainup as well.
So I think it's as much onthose of us who had been in the
industry for a while to helpcreate that pipeline and find
(33:14):
the curious people, cause I knowthey're out there.
I know there's people that arehungry to get jobs and learn a
skill.
We, as the more senior people,need to be willing to help them
along that pipeline as much aswe want to help ourselves.
Speaker 11 (33:29):
So the things that I've been stringing together
here, like there are differentemployers Expecting a fresh new
college grad to know it all andbe ready for everything that's
never been realistic, right.
And like there are legendaryprograms like Sprint had a
program called the Associate SCEprogram where you got hired and
(33:52):
then you spent 12 weeks taking12 week one week long classes,
testing and you got to pick yourassignment based on how well
you did in the program.
Cisco's had a great associatese program.
Juniper has gone up and downwith it, um, and I participated
(34:12):
in that right, um, and those areall good things, but none of
them are enough and I want toget this on the record here.
I think USNUA could make this anemphasis.
How do we do outreach to newand career people and
non-traditional?
I think that's another channelthat we've just not developed.
(34:34):
We all know so many people whogot into this without a college
education, right.
And the things that you don'tneed a bachelor's to do help
desk and to do so many otherthings right.
I think we can do more thingstogether.
I think if we could find a waywith US and UA and confederate
(34:56):
with NAF and NANAG and maybe putsome other industry programs in
place.
I don't mean this as a rebuke,but I'm tired of identifying the
problem.
I would like to start doingsomething about it, so I'll stop
there.
Speaker 1 (35:14):
I just wanted to wrap that part of it up.
My concern is that thenetworking vendors are going to
monetize the lack of peoplecoming in.
Their automation is going toget so good with their little AI
assistance.
And I don't I shouldn't havesaid little because, no, I'll be
honest with you.
I was just doing some stuff inSR Linux and they had this.
I was talking in the CLI innatural language, and it gave me
(35:35):
exactly what I needed.
That was a very differentexperience for me as a 15-year
network person.
So I see the advancement and Ithink it's much more likely for
a vendor to figure it out andhave one engineer do the work of
10 because their software isgetting so good, as opposed to
how we're going to get morepeople in.
We're going to go on a nationaltour and talk to high schools.
I think it's going to be hard.
(35:55):
Like people don't want to beplumbers, I think trade schools
are down too, so we're theplumbers of the internet, right?
So my crystal ball is vendorsare just going to monetize it
and we're actually not going tohave to plug that gap because
you're going to do the work of10 in a couple of years, because
the software is going to getthat good?
I don't know.
Speaker 3 (36:14):
Well, this is already happening in the startup space,
though, like startups areexpected with these AI agents
Okay, you have basically theknowledge of a few PhDs.
So it's shrinking the size andthe staffing of some of the
newer startups that are comingto market, and they're expected
to do way more with less,because you know that means
everybody gets more, you know,more capital at the end of the
(36:37):
run because there's not as manypeople involved.
Speaker 12 (36:39):
So I think it's all very good points about
educations and different ways tocome into the networking career
field.
When I first got into thenetworking career field, the
idea of having a Bachelor ofScience, even in being a network
engineer or even a seniornetwork engineer, was a foreign
concept for a lot of individuals.
I'd like to see us move back inthat path.
A lot of the network engineersthat I know came up from that
(37:01):
path.
They were found playing lasertag.
They were working in a lasertag store or they were working
at a gas station and they wantedto improve their life.
It was a career path to them,for you know a better way.
The thing with the AI commentand how AI is going to generate
all the commands that you need Ithink the one piece that I've
learned along the way is therestill is going to have to be
(37:24):
somebody who understands thatunderlying protocol.
I remember implementing my firstVersa SD-WAN solution when it
was on 15.2 and they had noflows or anything else to go
about it and I sat down withtheir engineers to figure it out
.
We were doing all those buzzterms to make the product better
week after week, no flows oranything else to go about it and
I sat down with their engineersto figure it out.
We're CI CD.
We were doing all those buzzterms to make the product better
.
You know week after week, butyou still had to be able to go
(37:46):
in there and understand what wasgoing on at the packet level
and how it was going out.
It was being deconstructed,being put back together.
That is where the networkengineer is going to excel in
the future.
If they're not stillunderstanding the history and
what's actually going on at theframe packet level and even
lower, we might not see as manynetwork engineers, but the level
(38:10):
of expectation of knowledgeyou're going to need is going to
be much higher, in my opinion,to be an efficient network
engineer.
Maybe.
Speaker 1 (38:16):
For the same salary?
Speaker 12 (38:17):
Yeah, for the same salary yeah, for the same salary
, it's not going to change muchyou're going to do four jobs for
the same, right.
Speaker 1 (38:21):
I mean, that was part of my resistance when all this
stuff came like I got I got tobe a network engineer, I got to
be an automation engineer, I gotto be a cloud engineer.
I got like really one salaryfor all.
But that's just, I think.
Speaker 12 (38:32):
But that's what's happening, right, that's the job
we saw that in otherengineering disciplines as well.
Think about the 1950s.
You saw drafters.
It took an entire room, or tworooms, of drafters to come up
with an aircraft.
Now you have what?
Three dudes sit down in frontof a computer and bam, now you
produce the next Boeing 777.
Speaker 8 (38:50):
This kind of dovetailing off what he said.
I mean, I think the largerquestion is how do you become a
professional?
How do you become aprofessional at anything?
Forget networks, right?
I mean, if you want to be arocket engineer, you don't go
study rockets, I mean, you studyphysics.
And how do you study physics?
You study math, and math is thebase principle of putting a
rocket.
If you want to go into themilitary, if you want to figure
(39:12):
out how war works, you go studywars from hundreds of years ago.
Right, I mean, there are baseprinciples of logistics and
things that go into codifyingmilitary science, right.
So, like, if you want to be aprofessional at anything, you
have to know what the baseprinciples of, and ultimately
the base principles of networksare electrical engineering,
algorithms, distributed systems,computer science, everything
(39:34):
else on top of it, the cli andhow you automate it.
I I mean, that's all likerecent stuff.
But if you want to, like he wassaying, like, fundamentally
understand or be a professionalat this discipline, go study
math, you know, and you'll beable to apply it over and over
and over.
I'm a, you know.
I was telling at the break,like you know, I was kind of
(39:55):
foolish enough to get two CCIEsback in the day.
But honestly, like, theundergrad in computer science is
worth so much more and amaster's degree is worth so much
more.
It teaches you how to think andif you know how to think, you
can apply it over and over andover, like just knowing.
Cisco CLI is a fad.
Speaker 5 (40:13):
I think about how young the tech industry is as a
whole.
I mean, you know, plumbers havebeen around for hundreds of
years.
That's a trade that's beenaround for a little while.
Think about all the otherindustries out there that are
hundreds or even older.
Tech is what?
Like 70 years old if we want tobe generous, 40 years old if we
want to be slightly lessgenerous.
(40:33):
So if that's how long we'vebeen doing it, there's still a
lot more to learn.
We're in the infancy oftechnology and I think you're
right to a certain degree.
Just having fundamentals thatare beyond technology and going
into concepts that never get old, like math basically, and being
(40:55):
well-versed in that, willguarantee that you can find a
spot somewhere.
But I don't think everybodywants to do that.
I think you're going to havepeople who are comfortable being
generalists and for thelisteners I pointed at myself.
I came from help desk, I wentinto sysadmin, but I worked at a
(41:16):
small company, so we wereresponsible for everything.
When I said everything, if theair conditioner broke, I had to
call the repair guy to come outand fix the air conditioner.
If the vending machine wasempty, I had the keys to go open
the vending machine and so,coming from that, I had to
understand how a storage arrayworked, how networking worked,
how VMware worked and also someweird stuff about Microsoft Word
(41:40):
that I don't even want to getinto.
But I had to have that sort ofgeneral knowledge and I was
comfortable with that.
I've never been a go deep onone thing and just be an
absolute expert in that onething and that's just.
I know that's me, that's theway I am, and I think that's
kind of where things are headed.
You can either be a generalistin the cloud or a generalist
with technologies, or you can gosuper deep on one thing.
(42:03):
Vendors are going to eat outthe middle, they're going to go.
We can make it good enough.
We can make a Meraki where it'sclose enough and easy enough to
click ops your way through it,that any small business or even
medium business never needs toopen a CLI because they can just
click through and connect allthe cables and it just works.
And if you're a generalist,that's perfect.
(42:23):
And when something goes wrongthey can call the network
engineer at Meraki, who is thedeep expert, and they'll figure
out what's actually going on.
And I think that's the caseacross all technologies, not
just networking is that softmiddle where we needed someone
with those skills.
That might be shrinking and youeither have to go deep or wide.
Speaker 13 (42:43):
I've been working computers for a very long time.
When I got into it there was noGoogle, so it's funny seeing it
grow.
So it's like CLI, that's allthere was back then.
I've grown up through computers.
My son's now 25, working for apretty big VAR and he's kind of
learned similar stuff Also onthe network routes.
He's doing segmentation andit's funny seeing the
(43:04):
differences.
So where I grew up with CLI hewas the opposite.
He's kind of gone the gooeyroute because that's what
they've learned.
And as someone also who's hiredpeople out of college, what I'm
kind of seeing is the tools aregetting better for networking
and you're going to see a lot ofways to make what you used to
do back in the early 2000s andeven before that.
(43:25):
It takes a very long time.
Things can get quicker, butwhen it really breaks those
tools are not going to be asgood and that's where you kind
of need the people who do knowwhat they know.
The people that worked at thebottom of the OSI stack and up
the way up know that stuff a lotbetter than the people who
start at the top and work theirway down.
Speaker 6 (43:43):
Yeah, but that's something that comes with time
and experience.
Speaker 13 (43:45):
Yes, so yeah, and that's the thing.
So it's like when, when I hired, we hired a kid out of college
great worker, he is a juniorperson now sent him to send him
to training so he getsexperience, which is great, he
learns his stuff, but the bigthing is he's shadowing the
people that they they may havelearned the books, but they also
learned where it's important inthe real world, because the
(44:07):
books don't always tell thetruth in a way, but that that's
yeah.
That's where they have to learnis is the younger kids need to
work with the older ones thathave been around so they can
then see where you know itdoesn't always work the same way
.
Speaker 2 (44:21):
So we we have about two minutes left, Um, and then I
think we hit an hour, which youguys are the experts, but my
understanding is you go over anhour on a podcast.
People stop listening, that'swhat.
Speaker 9 (44:30):
I'm told anyway.
Speaker 2 (44:37):
So I want to be cognizant of that, but I did
want to pause, you know, ask forclosing thoughts just relative
to the conversation.
Right, Much like our past nugs,we got through a total of one
question today, which is great.
Just love to do that.
It's good because it means thateverybody's involved.
Right, the conversation'shappening and that's really what
we want when we're here.
But just want to ask forclosing thoughts before we kind
(44:58):
of move into more of a wrap-upand we kind of close the event
out here.
Speaker 6 (45:01):
Yeah, I'll start.
First, I'm excited that therewas so much back and forth, like
we didn't just want to come up,so I appreciate everybody
weighing in with theirexperiences.
That was great.
It's great to have more of aconversation.
Second, I'm excited about thefuture of networking.
I think there's a ton ofinteresting stuff.
So, ladoo, really fascinatingproblems to solve and I think if
we pitch it the right way, wecan bring in a new generation of
(45:22):
network engineers and catchtheir curiosity and give them
all these cool tools and letthem go play and do incredible
things.
So I'm rosy about the future ofthis discipline.
Speaker 1 (45:32):
It's a great career, right, I mean it's a great job.
Nobody likes change.
I saw a lot of change.
I didn't like it.
I cried, Nobody cared, Got laidoff, panicked, learned a bunch
of new stuff and now I'm in abetter spot.
You know what's different fromnow than my 2012 CCNA?
Learn Linux, learn automation,learn cloud.
I think if you can learn thefundamentals of networking and
(45:56):
those other things, I thinkyou're in really good shape.
Speaker 3 (45:58):
I feel compelled to make my closing thoughts about
AI.
So seriously, though I almostcame like with the AI stuff I
was what I'm seeing it's kind ofscary almost is one of the big
(46:20):
skills over the years that Ilearned from not having Googled
everything and having an answerimmediately available.
Was it kind of the gentleman'spoint back there?
It helped me think and workthrough problems like I had to
figure it out.
I couldn't just type into chatgbt how do I do this?
And get an answer immediately.
Because when you do that foreverything, you find that's how
(46:41):
you train your brain.
That is the training thatyou're taking in.
You're not actually building askill set and that's something
that I honestly kind of scaresme.
But I know that AI is going tobe a big part of the future and
I was just in New York and Idon't know if you all know who
John Capobianco is.
He is really bullish on AI andhe actually made me excited
because he sees the good andwhere it's going and he kind of
(47:02):
narrows into where the realvalue is and is super excited
about it.
So that's just a reminder to methat all these new technologies
about it.
So that's just a reminder to methat all these new technologies
, you can't discount any of them.
You really have to see wherethey're going to be practical.
But again, how you use that chatGPT, like I know that I could
go in and say, hey, I want youto write this code for me, but I
(47:24):
still make my it's hard, but Istill make myself.
Not do that.
I might give it some of my codeand say, hey, what is wrong
with this?
But don't give me the answer.
I don't want you to give meanything back and I can get some
feedback and then go and figureit out with less stuff.
And I know that's still kind ofcheating, but at the same time
I still want to keep sharp so Ican actually write things and I
(47:44):
know how it works and I know howeverything you know connects
together.
So yeah, just I guess myfeedback and closing argument is
you know there's a lot of stuffcoming our way and you only
have so much time.
You know, try to prioritize anddo what's best for you and your
role.
You know what you're doing withyour job, cause if you can
apply it practically, you knowit's going to stick better and
(48:07):
it's going to serve you betterlong-term.
Speaker 2 (48:09):
With that we'll.
We'll close it out there.
Again, appreciate everybody'stime.
Thank you for the panelists.
Popular Podcasts
Dateline NBC
Current and classic episodes, featuring compelling true-crime mysteries, powerful documentaries and in-depth investigations. Follow now to get the latest episodes of Dateline NBC completely free, or subscribe to Dateline Premium for ad-free listening and exclusive bonus content: DatelinePremium.com
Stuff You Should Know
If you've ever wanted to know about champagne, satanism, the Stonewall Uprising, chaos theory, LSD, El Nino, true crime and Rosa Parks, then look no further. Josh and Chuck have you covered.
Intentionally Disturbing
Join me on this podcast as I navigate the murky waters of human behavior, current events, and personal anecdotes through in-depth interviews with incredible people—all served with a generous helping of sarcasm and satire. After years as a forensic and clinical psychologist, I offer a unique interview style and a low tolerance for bullshit, quickly steering conversations toward depth and darkness. I honor the seriousness while also appreciating wit. I’m your guide through the twisted labyrinth of the human psyche, armed with dark humor and biting wit.