Webinar: Kubernetes Made Easy
[Music]
Welcome to our webinar on cetes Made
Easy for
isps my name is Vikram Gupta I am the
product manager for cetes here at
digital ocean and I’ll be your host for
today we have an excellent line of of
presenters on this topic that I’ll
introduce in a
minute before we though I would love to
hear from all of you can you please take
a moment to share where are you dialing
in
from we definitely want to make this
session interactive and appreciate all
the
engagement by the way I’m joining from
SOS in California from
USA uh please feel free to say hi in the
comments and and share where you are
joining
from and we’ll also give a minute for
folks to join in
I see a bunch of folks llo Yash bojan
forgive me if I’m pronouncing anything
incorrectly Paul Carlos from
Portugal Lucas from
Poland
wow uh we we really appreciate that you
took time to join and we we hope that
the content is going to be valuable to
you and we and you’re going to enjoy
this he
Mark we’ll just keep another minute and
then get started I see that folks are
joining in a lot of folks are joining in
all right thank you for being here and
thank you for introducing yourselves let
me now take a moment to introduce my
fellow
presenters uh first we have Alexander
polishuk from
un. uh we are super proud to have un. as
our
customer Alexander can you please
introduce
yourself yeah hi thanks
um hi everyone happy to share my
experience today about digital ocean
kubernets thank you uh and next we have
adalo Delorenzo from
pener I myself have been using poer for
a while and I love the
product uh we are excited to have poer
as our partner today at ALU can you
please introduce yourself yes hello ad
Doo Lorenzo sales engineering manager
for Rainer and we’ve been a
long-standing partner of digital ocean
actually we’re huge fans of do and um a
lot of our backend development and and
um all of the dev stuff that happens at
poer is based on on your platform so
yeah we we just love you
guys thank you and finally we have my
friend Jack pierce my colleague here at
digital o Jack over to
you thanks PM yeah so I’m Jack I’m basic
in the UK I work as a Solutions
architect um so today I’ll be doing a a
demonstration of uh KU a thanks
speaker awesome thank you uh now we know
that many of you are likely startups or
independent software
vendors while kubernetes is a lucrative
and highly flexible platform it can also
be difficult to manage and that’s the
topic of our Focus today
how digital o and kubernetes can help
simplify operations so you can focus on
your core
business before we get started uh let me
also remind you that you can submit your
questions uh using the chat throughout
the webinar uh we’ll try to answer those
in the comments and then also follow up
towards the end
so today we’ll we we’ll start with a
overview of uh overview of Dos itself
and then uh followed by a real world use
case of how unit. a is leveraging
DS and then uh Jack will dive into a
live demo giving you a firstand look at
how simple it is to operate
dos and then we’ll explore poer and how
it enables migration to to kubernetes do
kubernetes and finally we’ll go through
Q&A and then hopefully we’ll share some
of the resources that will help you in
getting started with your dos
Journey startups and isbs are unique in
the sense that they are often building
products and services for their end
customers which is is a very different
set of requirements than somebody who’s
building a product for their own
conjuntion and and that’s where isps
need certain kind of customization and
flexibility from the
platform uh today we have 80% or more of
top do customers who are an
ISP and we have thousands of these isps
running on digital
ocean for example snippits is a
interactive video
platform bright data is a industry
leading web data
platform Nitro Pac accelerates websites
like WordPress and other CMS
websites and there are many more in fact
kubernetes is being leveraged by isps
which are building a broad array of SAS
services like in AI data analytics in
e-commerce in blockchain in educational
technology uh advertisement uh and
security platforms
Etc one thing that I note here is isps
are using digital ocean today not only
for development and production but also
as a means for
distribution today we have
marketplace where they can host one
clicks and S
add-ons and then we have a newly
announced partner program or partner pod
that offers a discount of up to 25% for
ISP
partners and
we we we really we we really strongly
recommend that ISP use our platform not
only for just production but also for
partnering with us for a win-win
situation and we we definitely want to
meet the ISP where it matters
most okay before we get started with the
do overview I want to get an sense of
where is everyone in terms of your Cloud
native uh Journey uh you see a few
questions on this slide like uh are you
have you adopted kuet in
production uh how about using a mix of
kubernetes and virtual
machines uh and is any of you using
using
multicloud uh I’ll just pause for a few
seconds please please take a moment to
respond in the chat uh that really help
us to customize the content to your
needs I’ll just pause for a few seconds
in case if anyone has any uh any
feedback otherwise please keep sharing
later as
well okay I I see that a few responses
like
mix okay cool uh thank you let’s let’s
let’s keep going in that
case first let’s start with the big
picture what are the compute options
that are available to you if you look at
this slide on the very left you have
droplets droplets are versatile virtual
machines that give you the freedom of
self-management with a with a variety of
plant and then you also have gpus
available through paper
space on the extreme right you have you
have app platform and app platform is a
fully managed service you just bring
your code we’ll build it for you we’ll
deploy it we’ll monitor it we’ll manage
it uh so that you really don’t have to
worry about doing anything about the
management you just bring your code and
we run it for
you and then you have a switch spot in
the middle which is Dos now the Dos
represents more like a balanced approach
in the sense that you get the power and
flexibility of
kubernetes without the operational
overhead uh we take care of setting it
up for you managing it monitoring it
upgrading it but then you take care of
running all the applications managing
your applications and you can customize
it the way you see fit
and and and that’s why that’s why for
isps who really need certain control yet
who want a managed platform platform B
is a is a really sweet spot for
them now let’s try to dig a little bit
deeper into this like why why even use
kubernetes for isps right I mean what
problem does it solve while this slide
shows a lot of different type of use
cases I just want to pick I just want to
pick one or two of those let’s consider
scaling so we saw different type of iscs
that build on kubernetes let’s take one
example that of educational technology
or attech when you are building an
edtech company uh you have online
classes you have students who are enroll
enrolling the classes may see a
significant load during the exam time
and during daytime almost no load during
late night and then there are certain
classes who see which will see a lot of
demand like lot of demand whereas
certain type of classes we will see very
few students it’s very natural right uh
if you are adopting a platform like
kubernetes then you can really rely on
the horizontal pod autoscaler which is a
feature of kubernetes to scale up your
class or the application based on the
demand based on how many students
joining and
then if that starts to you know reach
your cluster capacity then the cluster
autoscaler will automatically kick in
and it will scale up your cluster and so
you really didn’t have to worry about
writing code or doing things manually or
writing scripts to scale your cluster
based on the demand kubernetes does it
for you like this kubernetes provide you
automation a huge cncf ecosystem bunch
of tooling for data operations and then
the declarative architecture which means
automatic restart
Etc and and I I just want to summarize
by saying that isps need certain type of
customization Automation and scale and
that’s why kubernetes is the Swit spot
for
them well now let’s talk about why
digital oan kubernetes for isvs when we
build our product uh even even as part
of our engineering uh life cycle we
always keep two things in mind one is
how can we make it simple to operate and
then how can we genuinely lower the cost
of ownership for our
customers want to give one example on
simple to operate if you have tried out
digital oan kubernetes you know that it
takes one command
or one click on the UI to create the
cluster and then you want storage you
want load balancer you want database all
those are available to us as kubernetes
resources it’s that
simple and then let’s talk about the
lower TCO again I’ll give just one
example in the interest of time besides
the free control plan uh which is very
helpful if you are going to run like
tens of
clusters let’s talk about the upgrade
you are in isv you have 50 node cluster
when you are upgrading we are actually
allocating 10 extra node by default 10
extra nodes by
default so that your cluster workloads
will never go down and then we are
upgrading 10 nodes at a
time and imagine this upgrade happening
once or twice a
month
uh what what I mean we don’t charge you
for the sege noes that we are create
that we creating you absolutely don’t
have to worry about that likewise if you
are doing thousands of autoscaling
events per month which is very
natural we don’t build you until the
node is ready and part of the cluster we
don’t start building when the node is
getting booted up no we start building
clock only after the node is ready so
like this everything that we do we keep
in mind about how can we really uh how
can we really uh optimize the cost for
you
okay so so so far what we looked at was
why kubernetes for isbs and then why
digital Lo and kubernetes now let’s talk
about how isvs adopt
kubernetes I saw that many of you are
already running kubernetes some of you
are running a mix of kubernetes and
virtual
machines you have already figured out a
process right and so when you look at
the list of items here uh naturally even
though it looks like a lot it’s not that
overwhelming not really it’s a journey
so you basically you you basically start
with going from p p to production
quickly and then you think of
optimizing in Discovery you are looking
at uh okay what how b a cluster what
type of nodes am I going to need a load
balancer things like that and then in
development you’re looking at okay how
do you really go from writing code to
deploying in kubernetes without having
to
go through a lot of steps like how do
you make that iteration really really
fast cicd is a journey and what I see
that has worked for many customers is
they just start with a staging cluster
so that when they commit code it gets
automatically deployed to
staging whereas staging deployment is
automatic they use staging for testing
but the production is always always
manual until you really achieve that
level of confidence where you can
automate it so again cic Ser is a
journey where like you are slowly making
it incremental slowly making it improved
in production you are primar looking at
monitoring and scaling and primarily
monitoring security load balancer
configuration
Etc let me talk about scale
briefly uh it’s a good idea not to think
about scale until you really have a
stable production I’ve worked with a lot
of isps who go from PC to production in
a matter of day Zero to four to six
weeks and after that let’s say they
going from do0 Bill to $10,000 bill per
month after that I see them coming back
to6 or
$7,000 why that’s when they are looking
at scale their configuration what areas
can they optimize are they choosing the
right kind of nodes like are they using
shared or dedicated CPU memory
optimized by look at all those things
they are able to optimize the cost so
just to
summarize really while this shows a
comprehensive list the idea is to go
from zero to production in matter of
four to six weeks and then think about
scale after you have a stable
production uh that was a quick overview
of Dos uh with that let me hand over to
Alexander uh for for how how the LEL
un. yeah thanks Pam for the basics and
insights about digital ocean kubernetes
Services it was really helpful and
insightful but um before I go and share
my experience with digital ocean
kubernetes I will I would like to share
a little bit about unet as I said uh I’m
Alexander polishuk the C of unet and
we’ve built the product um based on
those three fundamentals create
communicate and collaborate let me
elaborate on those so first of all we
let you any user or any employees in the
company can come to our platform and
create AI model in minutes without any
technicalities and any technical
knowledges once you do that uh you would
be able to tailor it to your affirmation
and to your task daily task and communic
to communicate to it you can communicate
with your model through platform chat or
you can take it to your uh common
interfaces like email client CRM or any
other web applications and once you’re
happy with results what you can do you
can collaborate you can share that
access to your AI model with your
colleagues within the organization or to
anyone else literally anyone else so let
me give you an example the brief example
how it would be applicable for the
business use
case um so as you can see on my slide um
there is an example um each Corporation
or company has two valuable
assets first one is the business
information that the company holds and
the second was the second ones is
employees so imagine you can create a
variety of those AI models inside of
Viet um put the business
information aside the UH
responsibilities and then share it
within organization so all your
employees have access to those insights
at literally any minute of their
workflow they’re they’re not supposed to
go back and forth with the mail
communication now they’re able to access
information easily and the information
is really uh really good so now let’s
move along to the my experience with
digital ocean kubernets so how can how
did I came uh and decide to use uh
kubernets specifically with digital oan
so it’s a long story first of all I had
a long experience and I’ve been a long
customers of a digital ocean I’ve been
using lots of uh services that provide
but when it comes to creating a new
platform
it always matters how cost effective it
is especially in startup area so before
I go and uh create new system
infrastructure for new MVP or new
platform uh I always leverage those
metrics and I compare different vendors
and for some reason digital ocean always
comes on top um the the the cost is
significantly significantly cheaper
comparing to to other
vendors um then um when it comes to
deployment and development experience
it’s really simple from both sides from
the user interface SES and from the
documentation sites um my development te
team is really happy with the experience
they get with digital ocean especially
the wops team um they always um uh can
locate the information that they
actually need uh um within minutes
because the documentation is a piece of
art with the with the digital ocean
comparing to other vendors because I
have that experience coming to other
documentation and it’s easier to break a
leg
there um yeah moving along I would like
to share my experience um more technical
stuff about digital ocean what exactly
we’re using at unet and uh how it
benefits us basically I would say we use
the whole suit of services digital ocean
provide for our operation and the the
core piece of it is actually kubernetes
we use kubernetes as for staging and for
production environments and um the main
things I love about it is that digital
ocean Services they do not lock you up
with um their own services like other
vendors do and the second thing is that
it’s really reliable and when never had
an issue as bam mentioned before they
had that searge update they allocate
those notes extra notes for free of
charge and it we don’t we didn’t even
think about it but now I know I didn’t
know that insights uh about it about my
kubernets but now I know and I love in
it uh because I’ve never been charged
for it um the next thing uh is that it’s
pretty much easy to use as I mentioned
before um digital
um made a lot of work on
highlighting Essentials what you
actually would need to create your first
kubernetes it would take you literally
couple of days sometimes if you’re
experienced user of kubernets to deploy
your first MVP and then scale it up to
your needs and then do optimization
techniques Etc I I think the next uh
speaker will let you know uh give you a
small demo session and showcase how it
is easy to deploy
one um uh how about everything else um
we’re we are also using manage databases
object storages container registry for
our cicd load balancers and I I got a
mention that it was so easy to integrate
SSL certification with the load balancer
services and connect to my kubernetes
and when it comes to scaling as also
bigram already mentioned they have that
after scaler that really um simplifi the
work especially when it comes to
unstable traffic and most of the
startups have um let’s say high
availability traffic within the weekdays
and then on the weekend weekends there’s
nothing to do so my cluster go down and
up uh based with with help of the
digital ocean hype uh h PA
model um that would be it for me as a
technical um insights and yeah guys we
really love digital ocean as our
technical providers and we we’re looking
forward to use you for any new MVPs and
um development we we do thank
you uh I’m going to um introduce Jack
Pierce he’s going to give you a brief
brief uh demo session as I mentioned
before how to deploy one
kubernets architect for digital ocean hi
everybody my name is Jack Pierce and I
am based in the UK I work as a Solutions
architect for digital ocean so today I
want to give you a a brief demo of our
managed kubernetes what it’s like to
create a cluster what it’s like to
manage some of the settings on that
cluster create a low balancer and use
our block storage as well so I just want
to full transparency I will be skipping
ahead some parts of this recording just
because I’ve only got 15 minutes but I
will be available live on the webinar to
answer any questions afterwards I’d just
like to start by showing you this
reference diagram so here we have a
kubernetes cluster we have our Master
nodes over here this is our control
plane so we manage the control plane for
you that’s the hard bit you don’t need
to worry about that that is fully
managed by us low balancer as a service
can be integrated as well so that’s your
Ingress these low balancers can be
scaled to handle millions of requests
per second
we have our worker nodes so we have
pools of worker nodes these are made up
using our droplets so our virtual
machines and again these are all managed
for you so you you don’t get SSH access
or anything like that to these virtual
machines we manage all the difficult
bits for you the operating system the
file system all of the application and
software that are installed on those
virtual machines that’s all managed by
us so you don’t need to worry about any
of that so we have cluster autoscaler
and horizontal pod Auto scaling as well
built into the product for Block storage
you can use our volumes you can use our
S3 compatible spaces which is our object
storage or you could set up file storage
on a drop flip in the same private
Network as your kubernetes cluster
something like NFS okay so now to show
you what it looks like to create a
kubernetes cluster in the user interface
in the digital ocean control panels so
first we need to choose the data center
region the pricing for all of our
regions is exactly the same so you don’t
need to worry about different pricing
for different regions I’m going to
select London I’m going to select select
a VPC which is our virtual private
Network and any re resource that you put
into uh the same VPC will be assigned
automatically an IP address in private
IP space and those resources can
communicate with one another so any
droplets that you already have for
example in a VPC you can then put your
kubernetes cluster into that same VPC
and they’ll be to be able to communicate
with each other same with managed
databases as
well so we support the last three minor
versions of the Upstream version from
kubernetes so I’m going to deploy the
latest version here next we will choose
our cluster capacity so so we need to
choose a scaling type we fixed size so
we’ll choose the size of our nodes and
the number of nodes and that size won’t
change unless we come in and manually
update that this is good for predictable
workloads where you know what the CPU
and memory usage is likely to be we can
also then set an
autoscaling type Now with an autoscaling
type we set a minimum and maximum number
of nodes and this will allow us to have
a pool of servers that Scale based on
the load so if your application gets an
influx of traffic or if your application
all of a sudden has to process uh a ton
of uh workloads concurrently and it
needs additional CPU and memory to do
that the cluster autoscaler will notice
that it needs additional CPU and memory
and it will scale the number of servers
automatically within a range that you
set
so here I’m just going to set a minimum
of two a maximum of 10 nodes these nodes
are going to have 2 GB of memory I’m
just choosing the basic machine type
here or droplet the nodes are are
essentially just droplets and we can
choose from our premium range or our
dedicated CPU range we have CPU
optimized general purpose memory
optimized storage optimized they come
with different r ratios of CPU to memory
some come with 10 GB networking some
come with the latest generation Intel or
AMD CPUs so this is all explained in our
documentation you are given upfront a
cost for your nodes and if we were to
select a fixed size we’ll get a fixed
cost if we select an autoscaling nopol
we’ll get a a price range and it’s just
important to mention that we don’t
charge any extra for the control plane
we only charge for the nodes and we only
charge for any bandwidth overages that
you use so you’ll also get a very
generous bandwidth allowance with your
with your nodes as well so we have some
additional options here High
availability and database
operator when I enable High availability
it gives me a
99.95% SL
uh on my control plane so we will create
a replica of your control plane and that
will help maximize the availability and
uptime of that control plane so again
this is something uh just to give you
increased peace of mind um we also have
our database operator I won’t go into
that now but it’s essentially a way to
create and automatically link our
managed databases to your kubernetes
cluster um so we’ve uh we got to the
last stage and um I’m going to create
that cluster you you’re given upfront
costs for this cluster again very easy
very transparent to
follow okay so I’ve skipped ahead here
the cluster is now created so we get
this getting started and we can
literally just copy this command this
will use our digital ocean CLI tool doc
tool to save the kubernetes config
and it’s as simple as that so we’ve now
got access to our kubernetes cluster and
we can start deploying our workloads
just to note this is vanilla kubernetes
um you get full administrative API
access if you were to build a ton of
manifests and deploy them and get them
up and running and then you decide you
want to move that to um another Cloud
for Disaster Recovery purposes you can
do that there’s no vendor lock in
with our managed kubernetes so I just
want to show you how easy it is to
expose a service using a low balancer
our low balancers can scale to support
millions of requests per second so I’m
going to create an engine X pod
and like
that and then I’m going to
expose that engine X
pod using a
type of low balancer that will create a
low balancer on the digital ocean
platform so we can see that creating
here under external IP it says pending
that’s because the low balancer is
creating so if I come into the low
balancer section of the control panel I
can see that being created here now as
soon as that is ready it will be given
an IP address okay so that low balancer
has created and we’ve got an IP address
here so I’m going to curl this IP
address and I get a response from that
engine X pod so it’s as easy as that to
get your application your workload up
and running and then exposed behind a
low balancer so the next thing I want to
show you is how easy it is to consume
our block storage so if we uh go back to
our
kubernetes page we have this example
documented on the kubernetes page in the
control panel and it gives you an
example of how to consume our block
storage so to consume our block storage
we will essentially use a persistent
volume claim if you’re already using
kubernetes then you’ll be familiar with
this and you can you’ll probably
consuming you know persistent volume
claims using something like ice scuzzy
San but
it’s the same the same method we’re just
uh using a storage class that’s specific
to digital otion so if you have existing
manifests that are consuming your your
current block storage provider you can
literally just change the the storage
class name and we’ll we’ll be able to
create block storage volumes
so I have this example copied here in
this file so in this manifest I’m
creating a container with an image of
busy box I’m mounting a path of SL dat
and I’m creating a persistent volume
that’s 10 GB in size so I’m going to I’m
going to create that
[Music]
now and then I’m going
to retrieve my persistent volumes and
you can see that persistent volume has
been created and it is bound so if I
browse to my volumes in the digital tion
control
panel I will see that a 10 GB volume has
been created for me it’s been
automatically attached to this worker
node which is a kubernetes worker node
and yeah I can start consuming that
storage so the other thing I want to do
is just talk you through some of the
settings and configurations for managing
your kubernetes cluster so if you
already have a kubernetes cluster at the
moment on bare metal you’ll appreciate
that managing the control plane isn’t
Fun updating the control plane updating
the kubernetes versions isn’t Fun and
very time consuming especially when it
when you just want to focus on
development so some some options that we
have in here just to make your life
easier we have a digital ocean container
registry it’s a separate product it’s a
private container registry you can use
it to store your your images of your
docket images for example and if we want
to integrate that with our kubernetes
cluster we just come in here tick this
box will automatically create a secret
in your kubernetes cluster and then you
can start pulling images down from that
container
registry patching we can automatically
install minor patch versions that you’ll
be notified 7 Days in advanc and you can
come in here and you can select a 4-Hour
maintenance window as well that that’s
the time window that we will apply those
patches we also have surge upgrades so
you can enable this feature for free we
will add 10 additional nodes to your
cluster these are the upgraded nodes of
the new kubernetes version that you’re
upgrading to will move your will then
drain your existing listing nodes and we
will delete those those old nodes so
this is a way of upgrading your cluster
without having to compromise on the
stability or the availability of your
application we also have our kubernetes
Marketplace so you may have used our
Marketplace in the past to deploy
droplets templates for for droplets
which is really easy way of getting
certain applications up and running
we have a Marketplace for kubernetes
it’s oneclick apps it’s essentially Helm
in in the background so these are these
are just Helm charts and if we browse to
the marketplace the full marketplace
here we can see that there are 48
applications or preconfigured Stacks in
the
kubernetes category and we’ve got
everything here so we’ve got Ingress
controllers we’ve got monitoring Stacks
so if to click install app on this
monitoring stack here I’ll get
Prometheus grafana alert manager
pre-installed ready to go I don’t have
to do any uh manual installs of that
myself same for
Argos we’ve got Argo CD here as an
example so I’m going to install this to
my
cluster and I’m going to click on
install and I’m going to go back here
look at my pods and I can see that uh a
deployment has been made in the Argo CD
name space and this is uh an Argo CD
deployment I’ve got some getting started
instructions here so you can see the
installation um has just finished so uh
I can really easily just port forward to
my application in the kubernetes cluster
and then I can access Argo
CD and it’s as it’s as simple as that to
install an application from our
Marketplace and then get it up and
running it’s really easy to create and
manage backups of your kubernetes
cluster as well in 2023 digital ocean
Acquired snapshooter and snapshooter is
a great way to back up any resource that
you have in your digital ocean account
so it integrates n ly with digital ocean
so I can come in here and I can click on
add-ons in the control panel and I can
single sign on into the snapshooter
website once I’m logged into my
snapshooter account I can really quickly
and easily create and maintain backups
of my kubernetes cluster I can choose
which name spaces I want to back up I
can connect it to my own storage any S3
compatible storage and yeah super easy
to Main maintain those
backups so that was a quick introduction
to our managed kubernetes and just to
show how easy it is to get started we or
I would strongly recommend you check out
our documentation so docs. digital
ocean.com there’s some really great
getting started guides here as an
example we have the docs operational
Readiness things like installing an
Ingress controller setting up some
observability and they’re all step by
step so really easy to follow everything
that you would need to get started this
one as well is great building and
deploying your first image to your first
cluster so if you have a containerized
workload at the moment you have it built
as a Docker image or you just want to
play with a kubernetes cluster and get
started even if you don’t have your
workload containerized say for example
you have a PHP applic or a python
application this guide will talk you
through building it as a Docker image
pushing it to our digital ocean
container registry and then deploying
that as a workload on a kubernetes
cluster and you can follow this getting
started guide in about 15 20 minutes and
and you’ll have your application up and
running on a kubernetes cluster thanks
for watching this demonstration I will
be on the webinar to answer any
questions thanks everybody
hi
everybody okay hello I guess it’s me now
um well uh first of all thank you very
much all of you for joining and um I
also would like to thank digital ocean
for inviting uh myself and poer to be um
the speaker here uh we are huge fans of
digital ocean we’ve been um actually
long longstanding partner of dig digital
ocean and we’ve um we’re actually in the
de Marketplace you can deploy a poer
Community Edition straight into your
toets using um the
marketplace kubernetes cluster and run
poro just straight from there um so it’s
it’s quite um it’s been there for some
time and ourselves in our development
cycle we use do uh um a lot for um a lot
of the the internal Dev we do for painer
itself um so yeah we’ve been we’ve been
here for um for some time together um
working with this uh great platform uh
as everybody already said before very
straightforward to use um cost effective
um so yeah it’s what can I say I’m just
a huge fan basically right um so as you
see in the slide you can just go into
the deal Marketplace look for Port
tainer Community Edition and you’ll find
it there um I will I’m not really much
of a slide person I do have a second
slide following this one that sort of
shows what I’m going to do then I’m just
going to go straight and jump into the
demo um basically the idea is to show
how you can very easily migrate your
workloads and your deployments from any
kubernetes um uh instance running and
managed by poer into docs um I’m going
to use the Min iio um uh S3 bucket
server as an example of how to do that
uh and if you have any questions please
feel free to just jump in and ask them
while I’m presenting I’ll do my best to
answer those right so let me just share
my
screen uh let me select this one there
you go it should
be this one I hope
right I hope you can all see my screen
um well first of all yes I am a huge fan
I’d like to start with um complimenting
what Alexander said huge fan of digital
oceans tutorials I use them almost on a
daily basis for loads of things that I
try to use and and and my daily day um
adventures with containers or with Linux
or with um other operating systems so
they have a Hu really uh like to express
a huge thank you for for having this uh
huge knowledge base just available for
everyone it’s uh it’s amazing so I also
invite anyone to look into their
Community tutorials and their knowledge
base is just extensive and always up to
date and it’s saved my my skin several
times um here in the digital ocean
Marketplace I hope you can see it I
think I have to change the screen to
something else let me present just to
Ocean
Marketplace I’ll try to just share my
entire screen I hope this works
yep and try not to go into the endless
loop can you all see
my um screen I hope
so nope let me stop and start
again digital ocean
stream
one
okay all right so here in the
marketplace which I hope you’re seen if
you type in
painer you can just see from here that
you have the Community Edition and
variously deploy this as an app onto um
cuberes cluster very easily very quickly
it um Works beautifully so again we’ve
been in their Marketplace for some time
now and um and U we we actually very
happy about this partnership now let’s
go into the demo itself but if I change
my screen before I change it here for
some reason it’s not I’m sorry I’m going
to have to stop and start again and now
I’m going to share my screen with 14
there you go now you will see my screen
okay you should be seeing my poer
environment all right then so here I
have um three um kubernetes clusters one
is running on another cloud provider uh
the second one is running on bare metal
it’s actually running on a data center
that I set up as bare metal um and I’m I
managed this cluster via u a VPN so you
can see that the IP address is uh is a
um internal private VP uh IP address
this is running on another cloud
provider and uh the last one is running
on um digital ocean it’s it’s a docs q&
a cluster so here what I’m going to do
as I show showed in the slide I’m going
to
migrate this application here a Mini iio
Server
running here and I hope this
works yes it’s working um running here
on this uh bare metal kubernetes cluster
to my docks kubernetes cluster it’s
running a bit slow but um it should show
up very very soon hopefully
and it’s actually um I I really like Min
I hope you guys also had the chance to
enjoy this um this technology it’s
basically a dropin replacement for an S3
bucket running on on Amazon there you go
I already have a bucket set up called do
webinar and there’s nothing in it um and
it’s running again on my bare metal
kubernetes cluster so how would I
migrate this to um docs so as you saw on
the slide one of the methods that we uh
suggest is using git so we do have a git
integration so what I can do is just I
point to a uh deployment here and I with
this deployment I can very easily
migrate my deployment to docs so let’s
select docs here go into
applications and again uh doc be vanilla
uh um Cub cluster makes it a lot easier
because I don’t need to worry about any
specific custom uh configurations to
access this cluster I just can use it as
any of the cuines cluster and that’s
another thing that we really like about
do all right so I’ll just point I’m
going to select my name space um give it
a name going to call it I3 server for
example copy the repository here but
copy the
link um Porta does support
authentication if it’s the case and this
is any git repository it doesn’t have to
be specifically GitHub any git compliant
repository and I’m going to select Mino
right there it is with G Ops and there
are several options Po and web hook
basically I’m just going to use web hook
for the sake of this um the demo and I’m
just going to deploy it straight here
now once this is done I want to make
sure I can be able to access this and
the other uh lovely thing we we like a
lot about um T ocean is the ability to
just add um several U plugins as and and
applications onto my onto your cuber
cluster very easily I have already had
uh the um uh Ingress controller
pre-configured so just so we don’t waste
time because that was presented already
basically I have it’s running on this IP
address so what I’m going to do let’s
check it okay my Mini iio Server is
already running let’s expose this to
make sure that we can run it um from the
internet I’m going to add an
Ingress to
this me
IO deployment on my docs webin namespace
using engine X I don’t need any rewrite
annotations or um anything fancy for now
I’m just going to do a very
straightforward uh deployment
here the first thing I’m going to do
is um expose the console on the path I’m
add a new host I’m going to call this
guy mini iio and from here I am going to
expose the um S3 bucket service per se
so I have two hosts that I’m using with
a if you don’t know nipo nipo manages to
helps you map IP addresses to a host
name and with this I can very easily
access my mini iio deployment it should
show up here and if all goes well I
click on this link and I should see my
mini iio console there it is minio log
in
and there you go no buckets for now so
how would I migrate um my my data from
my internal one the one running on my
bare metal to the docks cluster very
easily I create a new
bucket call it do just the same webinar
2024 versioning and now I’m going to
replicate everything I have in this
bucket here running on my internal um
cluster to the one running on the on dos
basically what if we do this is very
easy just go here go replication I’m
going to add a replication
rule it’s um let’s copy the host name
again there you go it’s running AC in
Port 80 because it’s using the English
controller the password is P1 2 3 4 the
target bucket if follow as well it
should work webinar 2024 and I’ll leave
it synchronous for
now with existing
objects Oop There was an error maybe I
messed something up let’s
see um
save uh it should work oh sorry I forgot
it’s Mino Dot and without
TLS say this should work perfectly
lovely there you go now um as you see
there’s nothing in this bucket
uh if I want to synchronize things I can
very easily just go here and let’s
upload a file to see if it works upload
a file and you’ll see how Snappy this is
this is very very fast I’m going to
upload the logo here digital ocean logo
and if all goes well in this bucket I
should see something coming
up and there you go I managed to
synchronize um my internal data
center um um deployment into do and also
my data using minio inux obviously this
is just one method there are other ways
to do this you should be um aware that
there’s um data migration requires other
external tools like backup and restore
but as you saw it was very easy to do
and more so go very fast do is very very
very performant that’s another thing we
love about the platform and well that’s
that’s pretty much it that’s what I
wanted to show you guys how easy it is
to migrate using the get offs method uh
a deployment running on any um
kubernetes cluster in this case it was a
bare metal um cluster into docks by just
using a giops option and mapping the
ingresses very very easily from um my
internal deployment to the external
world all right then okay and that’s
that’s pretty much it it’s um we’ll stay
here for
questions um awesome thank you thank you
ad Alo uh Alexander and Jack now we are
ready to go to the questions
I did answer a few questions but I I
definitely see a few questions that that
we need to follow up uh if you have any
more questions please do chime in in in
in the comments right now uh so the
first question that I noted was from
JC uh J’s question is using kubernetes
in
aure uh the reporting can be overly
complex so how does the reporting in
docs compare to that and for example
resource allocation up time
Etc Alexander given that you are our
customer would you like to take a stab
at it sure um so when it comes to
monitoring um and comparing to other
vendors especially azour um I’m not
really um able to share my experience
with azour with Azure um that would be
better to ask my develop develops team
but uh when it comes to monitoring with
digital ocean we have it um really easy
to compare stats in the control plane uh
when it comes to high availability
traffic we can see the basics and we
also store the locks uh when it comes to
errors warnings and Etc security
vulnerability with um our inbuild system
uh that comes to our elastic seource
databases and we can see it through our
interfaces so uh I’m not really able to
compare it uh but yes when it comes to
Azure everything literally from my point
of view is really complicated I recently
had the that experience configuring the
applications there and I didn’t have fun
but let me be
honest good thank you uh and then the
the other question is from Calogero uh
my apologies I’m pronouncing it
incorrectly the question is can I have a
means of multicluster in
DOS and more specifically multi- region
and a Geo proximity service Jack do you
want to take a stab at
it yes sure okay
so I guess this is quite a common
question you can deploy kubernetes our
kubernetes clusters to all of the
regions that we have data centers um and
then you can install an overlay VPN so
like an overlay network using a VPN
technology such as tail scale or net
maker something like that that’s based
on wire guard um and then you can
communicate between the two clusters uh
you can um set up a service mesh that
way and and and uh communicate between
the two for
geo low balancing and Geo redundancy um
I would recommend taking a look at
something like um Cloud Flair’s geolo
balancer because that will integrate
nicely with um arow balances so then you
can fail over traffic between the the
two
regions awesome uh thank you thank you
Jack uh then the other question is from
quy Garcia can you elaborate a bit more
on the database operator for docs and
why would I want to use that
uh maybe I can take it up uh we build
this database operator so that you can
create a database managed database as a
kubernetes object so today basically you
have to go and create it using your own
terraform or CLI or UI and then use it
from kubernetes but what if you just
wanted to create the database itself as
a kubernetes resource that’s why that’s
how you can use the use the database
operator if that suits your automation
needs great if not you should just
create the database the way you want to
the way that you are creating
today hopefully that answers your
question and then from Max we had audit
logs uh I I already answered that in the
chat uh I don’t see any other question
and we just have two minutes remaining
so I think it’s time to just close it
here uh we did share some of the
resources here for you those those who
you want to like if you are new to
digital oan definitely sign up and then
you get $200 credit if you need help you
can talk to sales if you want to explore
what un. provides you for for really
being more productive at work look up
their
website uh and then for pner uh
obviously if you want to try it out you
have you get up to three notes for
free uh with that that brings us to the
end of today webinar uh I I really hope
that you found the discussion helpful uh
on behalf of the entire team I want to
thank you for your engaged
participation have a wonderful rest of
your day thank you
To learn more about DigitalOcean: https://www.digitalocean.com/
Follow us on Twitter: https://twitter.com/digitalocean
Like us on Facebook: https://www.facebook.com/DigitalOceanCloudHosting
Follow us on Instagram: https://www.instagram.com/thedigitalocean
We’re hiring: http://grnh.se/aicoph1