Hang tight while we fetch the video data and transcripts. This only takes a moment.
Connecting to YouTube player…
Fetching transcript data…
We’ll display the transcript, summary, and all view options as soon as everything loads.
Next steps
Loading transcript tools…
Streamlining Scalability: Mastering Aerospike with Kubernetes Operator | Aerospike Summit India 2024 | Aerospike | YouTubeToText
YouTube Transcript: Streamlining Scalability: Mastering Aerospike with Kubernetes Operator | Aerospike Summit India 2024
Skip watching entire videos - get the full transcript, search for keywords, and copy with one click.
Share:
Video Transcript
Video Summary
Summary
Core Theme
The Aerospike Kubernetes Operator (AKO) is presented as a solution to automate the deployment, management, and scaling of Aerospike enterprise clusters on Kubernetes, simplifying operations and reducing manual effort.
Key Points
Mind Map
Click to expand
Click to explore the full interactive mind map • Zoom, pan, and navigate
[Music]
Well ladies and gentlemen for giving us
the download on mastering aerospike with
Kubernetes operator. I invite now on
stage Miss Shri Nidhi Kashik principal consultant
consultant
So today's topic is for folks who have
already on aerospike or looking forward
to using cubernetes to manage aerospike
successfully. So this is the uh this is
all about that particular
topic. So first thing is a quick agenda
of what we are going to follow through.
This is a quick agenda. I'll just not go
through this for from a timing
standpoint. So what we have right now if
you're facing any of these issues or any
concerns related to running your
uh aerospect clusters manually unable to
deploy on time finding scaling issues
the amount of effort that you're uh
investing in upgrading is huge or you're
running or you're dependent on runbooks
provided either by yourself or developed
by aerospike or you're seeing something
like your Kubernetes this cluster is
being managed by centralized
decentralized team which in turn uh
causes something like low engineer
morale or non-adherance to standards. If
any of these is what you are facing
right now as of today here's this
subsequent slides which talks about how
we can elevate this problem. Now from a
consequence standpoint we have seen that
whenever these issues are being faced we
this leads to something like missed SLAs
increased cost in managing your home
ground op automation as well as missed
revenue opportunities because
you're investing more time in doing
mundane teams like upgrading or
maintaining your aerospike which should
be given out for more innovative uh
efforts. Now this brings up to the next
slide. What is what is AKO all about? So
AKO or Aerospike K operator helps you to
automate the deployment management of
Aerospike enterprise clusters on
Kubernetes. Then why Aerospike AKO?
Because this helps you to reduce the
resource consumption chances of error in
while you're managing these parameters
and make sure makes your S sur lives
easier. That is what AKO is here to help
you out with. Now this is a very simple
view of how it architecture looks like
wherein on your right hand side depicts
the user. User provides the inputs to
the CRD files and we have host of CRD
files in our repository which caters to
each and every use case that you have in
mind. Right? The example that we are
going to showcase is just for demo
standpoint but much more complex
requirements can be managed
and heavier aerospike instance can be
set up with minimal fuss. Now from a CRD
standpoint is more of an YAML file that
we have which calls out the
configurations associated with your uh
aerospike environment and we have this
Kubernetes aerospike operator which
interacts with your Kubernetes API and
it helps you to do the next activity
which is to help you to horizontally
scale vertically scale upgrade your
aerospike instances with no uh
interference from the user as well as to
try changing parameters also. right
which needs rolling restart of your
environment and all of this is happening
mind you without any sort of downtime
right that's the beauty of using AKO now
we are compatible on with Amazon EKS AKS
Azure Kubernetes services Google
Kubernetes services and if you have an
inos kubernetes setup we are compatible
with those as well the same yl file the
same yard stick holds good except the
parameters would change based on the
cubernetes in uh cubernetes setup that
you have at your
Now this is just a small example of how
uh configuration looks like and how
complex an aerospike environment would
look over here on your right hand side.
So we have an operator connecting to
your Kubernetes API. You have a option
to c capture your configuration maps.
Secrets like TLS files, your
configuration files can be kept under
this secret directory. In addition to
it, we can also have a aerospike
Prometheus exporter as a sidecar which
will help you to monitor these pods or
clusters effort effortlessly. And mind
you, this is a single pane of operating
window. We don't have to have multiple
operators. One operator manages the
entire fleet of clusters that you have
from a aerospike standpoint. That's the
beauty of
AKO. Now going to the demo. So the
intent of this particular demo is to
yeah let me let me give you a voice over
makes it much more interesting. So what
we are trying to do over here in this
particular case is we have set up a GKE
cluster on uh Google cloud. So on your
left hand window what we have done is as
you can see we have installed u uh
aerospec operator and this is a jump
server on which we have aerospec
installed and what we are trying to do
in this particular case is
to there's a sample of uh the repository
details that I was talking to you about
the github link which gives you these
particular parameters for example what
I've done is I've taken a yl file called
submit 6x6.7x wherein I am intending to
run something like set up a cluster
called a cluster with three nodes.
There's a three pods that you see under
the size and what version we are talking about
about
6.1.0.1 as a version and this also
contains a snippet of the storage class
that we use from a role-based access
control. We have the users that has to
be generated automatically and finally
the aerospect configuration on what sort
of in installation we want. In our case
what we have done is we have used
something called memory size a pure
inmemory cluster for just for an
example. Now what I've done is I've just
run the cubectl appli minus f command.
And if you can see here it says cluster
created. Please concentrate on your
right hand window. What you have done is
we have just tried to access the three
pods that we have set up is shown up
over here. Now these pods represents
each and every pod that is associated
with the aerospec cluster. The reason
being we had asked for three pods to be
set up over here. Now what we are trying
to do here is to access it and show you
that these environments can be accessed
using uh the ACDM command which acts
like a client. If you're able to client
use this client the same thing that uh
goes with the any other client whether
it's Java, Python, name it we have we
can connect to this particular
environment using these commands. So
this asadm gives you the value. The
admin user that we have set up as part
of your rolebased access as this control
gives me access to it and whatever we
have set up as a YAML file gets set up
as an environment. AO cluster is my
cluster name.
E6.1.0.1 is my cluster value and the
cluster size being three right now. So
this is just a first part of the story.
Now much more complex things. For
example, you want to do something like
an upgrade of an existing environment
and you don't want to spend most of the
time. So what we have done right now is
we are trying to upgrade this
environment from 6.1 to 7.1 as well as
to increase the size of the uh the
number of pods to four and we just go
ahead to apply this particular value and
that is what is being showcased as part
of this demo and how seamlessly uh this
gets migrated to the next upgraded to
the next version in a sequential fashion
or a rolling fashion is what you'll see
on a window on your right hand side.
Again, we have applied the cubectl apply
command with the new YAML file that we
had right now. And that takes you to the
next window. So, what I've done is I've
just copied the configuration to
increase the value. And that's what
you're seeing over there. And here's
where we are applying the same YAML file
that we had initially triggered. And if
you see there, it's showing us
configured. And that now we take you to
the right hand window wherein we are
doing something like an info network
which tells you what sort of build was
installed over there. Now if you can
concentrate on this particular column
over here called the build migration the
build E610 that's what we started off
but what we have put up right now is
7.1. So what's going to happen over here
now one by one you will see that each of
this node will be brought down we
upgrade it and we are glad to announce
that we support something like mixed
clusters wherein we work with both 6.x
and 7.ext at the same time. So without
any sort of downtime in a rolling
fashion we are able to migrate from a
older version to a newer version of uh
aerospike and that's what you're seeing.
The first one is already through. So it
waits for the migration everything that
you are expecting from a end user is
automated as a best practice within the
AKO. Now if the first one is already
done then the second one will follow
suit and that's what you see by running
the info network again and this takes
couple of seconds or or the time it
takes would be the time for the
migrations to complete. So you'll see
the second one going through right now for
for
migration. It's up. We are left with a
third one. So for for the time being we
have just two of them already migrated.
The third one will follow suit shortly
and the next step would be to increase
the number of parts. Initially we had
given it as three and now it'll get uh
migrated. It'll become four parts very
shortly. See this is the end result that
you wanted. So seamlessly uh without any
sort of manual intervention using the YL
inputs that a single place from where
you can run the entire upgrade process
and finally the fourth node as you can
see is being spun up. So this sort of
automation is what you get using AKO and
every other best practices that we have
has already been built into this
particular tool. So this is about the
demo. So first question that comes to
your mind is there are a lot of moving
parts right has this ever been tested by
yourself or by any other client of that
you know of. So we are glad to announce
that we have one of our part key
partners which is Flipkart who has been
through this journey and uh I think is
there folks from Flipkart who can raise
their hands
please. Yeah, great. These are the
wonderful champions like FI membered
superheroes who are managing somewhere
close to 300 odd clusters with zero
downtime and the latest sales day they
were able to uh manage somewhere like 35
billion transactions uh over over this
particular 300 cluster with zero
downtime and that's the beauty of how
simple is it and how effectively the
team over here has been able to manage
it as well as we have also have critio
as well which is another customer which
has almost the same range we have many
more customers who have done this right
now and aerospike has your back. So
ultimately we have our engineering team
which is all ears to see what how best
can we evolve this product further. So
we have got excellent interactions with
the individuals who are already using
AKO and we hear them to make sure that
the subsequent releases has those
So from a maturity model standpoint
where does operator stand right from a
maturity standpoint we are at level
three wherein we are working on creating
something like a backup failure and
recovery built into this particular AKO
and by end of this September you should
be able to see that these features will
be there in AKO number one as well as we
are moving our sites into the next one
wherein we want to do deep insights on
how do we metricize how do we read these
metrics create alerts, meaningful alerts
and make sure that we process these logs
which in turn will lead to something
like a autopilot mode wherein you want
to scale up much more effectively,
automate the entire flow, all those
things will be part of level
five. So that brings me to the next
slide which is the benefits of ATO.
After seeing this right now all these
the wonderful things that's moving
around you can make sure that you can
have a automated deployment scaling and
management much easier. That's number
one. managed by the centralized team
just like Ash for show showcased by our
colleagues over here as well as it leads
to higher engineering model because 300
400 odd clusters being managed by five
individuals is not a a joke right that
these guys have been able to do that and
it leads to a higher engineer morale and
adherance to standards being managed by
a centralized team brings in something
like an adherance because they know what
are the best practices that has to be
involved so they make sure that that is
adhered to and finally Every other thing
has got a positive business impact right
like effort that they save over here can
be channeled towards in innovative
efforts which in turn leads to something
like a dollar value SLA is being met
like what's being showcased and zero
downtime is what you will definitely get
from an aerospike which which is taken
over here so we are glad we have our
booth over here which with any other
questions that you have we'll be glad to
take it up and as I said this demo is
just a teaser there are much more
complex things that can be managed We
have a host of uh GitHub repository CRDs
which gives you entire combinations that
you can try out like XDRs making sure
that you want to change change some
parameters those are all available.
That's about it from my end. Thanks for
listening and have a wonderful day. [Music]
Click on any text or timestamp to jump to that moment in the video
Share:
Most transcripts ready in under 5 seconds
One-Click Copy125+ LanguagesSearch ContentJump to Timestamps
Paste YouTube URL
Enter any YouTube video link to get the full transcript
Transcript Extraction Form
Most transcripts ready in under 5 seconds
Get Our Chrome Extension
Get transcripts instantly without leaving YouTube. Install our Chrome extension for one-click access to any video's transcript directly on the watch page.
