►
From YouTube: Sept 30, 2021 - Ortelius Architecture Meeting AUS TZ
Description
No description was provided for this meeting.
If this is YOUR meeting, an easy way to fix this is to add a description to your video, wherever mtngs.io found it (probably YouTube).
B
Yeah
sure
so
we're
we're
both
based
out
of
australia,
I'm
currently
in
new
zealand
until
next
year
on
the
the
head
of
cloud
engineering
for
a
fintech
company.
So
we
yeah,
I'm
generally
always
staying
in
the
fintech
space
been
working
with
kubernetes
for
a
couple
years
now
and
you
know,
have
experience
over
all
the
cloud
providers,
except
for
alibaba,
actually
yeah,
very
similar
backgrounds:
okay,
okay
and
emma
you're,
pretty
much
the
same
right.
B
Okay,
sorry
updates
I'll
just
share
my
screen
very
soon.
All
right,
let
me
bring
up
the
argo,
so
there's
a
few
decisions
to
be
made
today.
So
we've
made
progress
on
the
mono
repo.
B
Now
the
decisions
we'll
probably
have
to
make
which
cluster
are
we
deploying
ortillius
to
because
we
obviously
have
the
small
devops
cluster,
but
that's
very
small,
and
I
don't
want
to
scale
that
up
for
no
reason,
so
this
is
actually
failing,
because
there's
not
enough
capacity
at
the
moment
so
steve
do
we
want
to
deploy
to
your
one
with
that?
Has
the
docs
and
stuff
on
it
or
do
we
want
to
start
off
by
deploying
to
this
cluster.
D
I
think
we
could
take
the
the
devops
one
and
just
make
that
one
a
little
bit
bigger
if
we
need
to
yeah
and
then
once
we
get
everything
proved
out,
then
we
can
go
and
hit
up
the
the
production
cluster.
B
Yeah,
I
also
I
forgot
his
name.
Sorry,
I've.
Also.
You
can
see
that
I've
deployed
captain
using
argo
for
the
captain.ortelista
io
okay,
so
that
way
adam
and
starts
with
us.
D
B
Working
on
the
captain
piece
with
adam,
so
that's
an
environment
making
use.
So
that's
in
the
fortillion
communities
reform.
So
that's
an
update
on
that
side.
We
have
I've,
got
that
going
through
from
as
your
front
door,
just
a
really
standard
cheap
one.
So
we
can
enable
multi-region
down
the
line.
If
we
wanted
to
start
the
control
planes,
you
know
the
more
advanced
stuff
than
we
could
do
and
then
yeah.
B
So
if
we
want
to
so
on
that
decision,
we
can
then
make
this
cluster
a
little
bit
bigger
and
then
start
deploying
everything.
So
if
we
go
to.
B
The
application
set
so
mono
and
poly
so
also,
maybe
you
want
to
talk
here,
but
so
for
all
the
micro
service
repositories,
we're
saying
anything
in
the
organization
that
starts
with
alternative.
Ms
is
going
to
be
we're
going
to
scoop
that
up
in
the
deploy
branch
in
the
folder
chart.
D
D
Yeah,
that's
what
you
scoped.
The
two
is
perfect.
We
have
a
bunch
of
other
like
demo
repositories
around
the
hipster
store,
that's
out
there
and
those
start
with
like
store
dash,
something
for
the
repo
name.
So
what
the
filter
that
you
put
on
there
is
is
the
correct
one.
B
Okay
and
then
I
believe,
one
of
the
one
of
the
charts
is
not
in
deploy
it's
in,
like
it's
under,
is
underscore
s
vc
for
service
or
something.
D
Yeah,
I
I
I'll
I'll
make
sure
that
the
branch
gets
moved
over
I'll
rename
that
branch
and
push
it
up.
B
Sure,
okay,
cool
awesome
cool
and
then
this
is
going
to
our
private
cluster
as
well,
so
that
that's
working
good.
A
So
I'm
just
trying
to
understand
so
brad
you're
trying
to
use
argo
cd
and
set
up
a
github's
process
for
microservices
of
artelias.
Is
that
correct,
yeah
into
a
micro
into
a
azure
eks
cluster.
B
This
one
here
is
called
captain
yep.
It's
the
same
thing:
yeah.
A
Okay,
kept
in
like
autelius
product
or.
B
B
Just
a
quick
run
through
like
this
is
the
the
front
door,
so
this
doesn't
cost
as
much
as
it
doesn't
cost
that
much
it's
just
a
standard
one,
but
I
did
the
ticket
to
the
cd
foundation
and
then
they
put
in
my
cna
records
and
then
we
can
add
more
if
we
want
it
to
go
to
the
anything
else
as
well.
But
we
don't
need
to
talk
too
much
about
this,
but
just
just
so
we
have
captain
and
aggressively
and
then
they
brought
into
the
ip.
B
B
Yeah
and
then
the
cool
thing
to
connect
to
this
is
we
can
connect
through
bastion
service
because
it's
a
private
cluster.
So
you
come
in
here
and
I
actually
spin
these
down
when
I'm
not
working
on
it
to
save
money,
because
fashion
service
costs
eight
dollars
a
day.
A
Is
there
a
document
or
anything
that
defines
the
basic
setup
or
is
it
something
that
we
need
to
talk
through
about.
A
A
Yeah
one
of
the
challenges
with
githubs
that
I
have
faced
brad
while
you're
connecting
to
it,
is
primarily
how
you
set
up
your
struct
project
structure
in
git
right,
because
every
company
has
a
different
way
of
looking
at
the
git
project.
A
The
most
recent
one
that
we
have
done
with
with
vworks
git
ops
tools
we
had
challenge.
We
had
to
use
multiple
registries
or
multiple
images
from
the
same
registry,
because
we,
the
approval
mechanism,
was
based
on
the
image
per
environment
which
which
ideally
is
not
a
good
good
practice.
I
would.
I
would
like
to
have
a
single
image
of
any
microservices
that
that
gets
pulled
into
each
environment,
because
it's
like
an
artifact
that
you
have
tested
in
a
lower
environment,
going
up
higher
up
right.
C
Absolutely
so
the
process
which
you're
explaining
is
basically
the
flow
we
are
following
is
your
first
environment
will
be
devastating
once
it
merges
to
master
it
shouldn't
really
matter.
Everything
basically
starts
going
through
into
all
your
respective
environments.
That's
where
captain
comes
into
for
to
control
the
quality
gates.
So
if,
for
example,
staging
is
approved,
then
staging
should
automatically
get
deployed
with
whatever
got
approved
via
captain,
and
it
will
just.
A
A
So
yeah,
so
so
that
is
where
a
little
bit
my
head
is
not
clear.
I
know
in
the
previous
discussion
with
tracy.
She
was
she
also
kind
of
explained
the
same
way
that
how
how
this
is
going
to
work,
but
does
that
mean
that
we
are
having
separate
repos
that
per
environment,
because,
typically
in
a
ci
process,
you
don't
have
a
merge
that
happens
for
environment
right?
You
just
push
your
artifact
from
one
environment
to
another
environment.
C
Yes,
but
the
environment
to
another
environment
is
more
infrastructure
based
and
then
whatever
state
is
controlled,
your
state
is
generally
controlled
in
your
respective
ci
cd
pipeline
or
kind
of
that
was
the
old
way
of
git
ops.
So
the
new
way
is
you
manage
it
via
multiple
things.
One
is
your
commit.
Ids
are
important
or
your,
for
example,
image
charts
are
important,
so,
for
example,
if
I
have
a
dev
environment
and
it's
on
branch
dev
first,
I
have
a
separate
application
set
for
dev.
A
Okay,
so
so
your
image
is
b
so
here
here
is
how
I
mapped
it
in
my
head,
based
on
what
you
said,
you
have
a
project,
let's
say
any
micro
service.
Let's
say
it's
an
email
service,
an
email
service
would
have
multiple
branches
like
dev,
staging
dev,
qa
and
staging
assuming
that
so,
whenever
you're
ready
for
dev,
your
dev
image
gets
built,
and
the
last
comment
id
in
that
branch
is
what
your
image
gets
tagged
with,
as
that
is
your
image.
C
C
A
Okay,
so
once
you
are,
let's
say
your
dev
testing
dev
is
validated,
then
you
merge,
dev
or
so
your
captain
will
say:
okay,
you're
now
you're
ready
for
qa
right.
So
what
happens
with
that?
At
that
point.
C
At
that
time
and
point
you
will
basically,
the
application
set
is
just
running,
as
is,
but
your
captain
then
goes
and
makes
an
update
on
the
image
on
the
value
for
that
respective
environment.
Okay,.
A
Okay,
so
your
your
image
shark
gets
updated
yeah.
So
that
means
that
branch
has.
That
branch
is
now
ready.
But
a
question
here
is
that
how
are
now
what
happens
to
the
docker
image
in
the
registry?
Are
you
going
to
update
that
as
well,
because
it's
a
pull
based
deployment
right
that
that
needs
to
be
updated
is
what
I
would
imagine.
A
Correct
so
so,
if
so,
your
captain
is
going
to
update
the
docker
tag
here
in
the
qa
branch
and
automatically
just
the
values.yaml
okay,
the
qra
values.yaml.
So
immediately
that
figures,
a
docker
build
in
the
ci
pipeline.
D
C
D
It
would
just
reuse
that
previous
image,
now,
if
you're,
if
you're
now,
one
of
the
things
is,
if
you're
going
from
from
dev
to
qa
and
you're,
going
from
dev
to
qa,
and
you
you're
doing
a
merge
and
you're
getting
a
new
commit
id,
because
that's
where
you're
pulling
all
your
code
together
and
that's
going
to
be
your
last
build,
then
you
would
at
qa
stage
you
would
get
a
new
image
tag
and
a
new
sha
and
that
would
be
used
in
the
values
file
going
forward.
D
It
just
depends
on
where,
in
your
pipeline,
your
last
build
is
going
to
be
of
the
image
most
people
at
the
ci
level.
D
D
We
have
a
new
show
and
then,
when
we
go
to
like
staging,
we
would
just
reuse
that
image
tag
and
sha
for
the
staging
and
that
at
that
point
we're
just
pulling
re-pulling
from
the
the
registry
we're
not
rebuilding
at
that
level.
Then,
when
we
go
into
production,
we
just
reuse
that
the
same
image
again
so
once
we
get
to
the
higher
stages,
we're
going
to
be
reusing,
the
same
same
image
of
what
we
tested.
A
C
To
give
you
a
clearer
picture:
brad,
can
you
open
the
application
sets?
I
think
it's
more
descriptive
there.
Then
I
was
trying
to
explain
it.
Oh
yeah
sure
so,
scroll
down
to
the
part
where
we
were
picking
up
the
headshot.
C
C
To
have
it's
a
list
so
you're
going
to
have
another
list
where
production
and
you're
going
to
have
staging
and
you're
going
to
have
dev,
etc
and
line.
Number
39
is
the
one.
Well.
Currently
it's
commented
out,
but
once
you
put
in
the
production-
and
you
have
multiple
values-
file
you're
going
to
have
it
cluster
hyphen
values.yaml,
where
cluster
signifies
the
environment,
it
isn't,
and
you
should
ideally
have
a
template.
Values.Yaml
file
with
one
environment
variable
which
needs
to
be
replaced
every
time,
there's
a
new
image
which
gets
built
on
the
dev
level.
A
Okay
got
it
so.
A
Okay,
so
so
how
my
I
just
need
clarification
on
how
you're
updating
from
dev
to
staging
what
is
like
what
is
the
process
there.
D
So
that
is
going
to
be
where
the
interaction
between
artelias
and
captain
will
come
into
play,
but
at
the
end
of
the
day,
what's
gonna
happen
is
there's.
Like
ahmed
said,
was
there
there's
gonna
be
a
commit
with
the
well
from
from
dev
to
staging?
Do
you?
Let
me
ask
this
question:
do
you
want
to
rebuild
the
image,
or
are
we
going
to
take
the
dev
image
and
move
it
out
to
staging.
D
For
well.
A
Yeah,
no,
I
I
so
I'm
trying
to
understand
your
process
right.
So
typically
you
you
can't
really
move
the
image
because
from
dev
to
staging,
you
need
to
update
the
image
tag
at
least
or
rebuild
the
image
or
update
the
image
tag
for
for
changing
the
the
registry.
D
Yeah,
so
what
would
happen
if
we're
going
to,
let's
say
we're
going
to
go
we'll
put
another
step
in
the
pipeline?
Just
make
it
easier.
So
we
have.
We
have
our
dev.
We
have
qa
staging
and
production.
D
So
when
we
go
from
from
dev
to
qa,
that
would
that
when
you
do
that
pr
you're
gonna
end
up
getting
a
new
commit
at
the
end
of
the
day,
with
the
for
for
the
at
the
qa
on
the
qa
branch,
for
example,
from
there
captain's
going
to
kick
in
and
recognize
that,
there's
a
new
commit
on
qa
and
it's
going
to
start
it's
a
its
pipeline
which
will
do
the
docker
build
the
tag.
The
push
so
there's.
A
D
At
the
dev
to
qa
level,
that's
where
the
the
pr
would
be
now
once
we
get
beyond
qa
and
because
captain
like
ahmet
was
saying:
has
these
quality
gates?
We
can
then,
and
the
ability
to
do
run,
test
cases
and
things
like
that.
We
would
then
go
ahead
and
use
captain's
approval
mechanism,
with
automated
approval
and
quality
gates
to
move
from
qa
to
staging
and
then
once
we're
happy
with
it
in
staging
it
would
automatically
move
to
production.
D
So
the
last
like
human
touch
would
be
the
pr
going
from
dev
to
qa.
E
D
And
like,
like,
I
said,
there's
approval
gates
on
the
captain
side.
So
if
you
wanted
to
wait
till
midnight
on
friday
to
do
the
deployment
into
production,
they
have
quality
gates
and
gated
pieces
in
in
captain
that
allows
you
to
set
that
type
of
scenario
up.
A
Okay
and
and
you
are
there
separate
registries
per
environment
or
that's
what
I
would
do
sorry.
C
D
Registration
yeah,
so
you
may,
you
may
have
a
different
tag,
so
you
may
have
the
the
dev
would
be
tagged
one
way
and
then
once
the
the
pr-
and
that
may
be
tagged
like
with
the
pr
that
you're
working
on
or
something
like
that
right.
A
Because
you're
updating
the
helm
chart
with
the
tag
per
environment
with
kept
with
captain,
I
think
a
single
registry
could
have
the
same
image
with
multiple
different
tags
and
and
the.
D
Yeah
now,
when
we
get
to
qa
and
we're
working
with
qa,
you
know
one
of
the
best
practices
is
not
to
use
tags
but
use
the
shot.
The
digest,
because
that's
immutable
and
gives
you
a
better
audit
trace
down
the
road.
So
when
we.
D
A
A
Okay,
I
think
yeah,
I'm
I'm
I'm
a
bit
clear
now
on
how
how
you're
doing
the
key
tops
yeah,
but
definitely,
I
think,
just
putting
together
a
high
level
diagram
could
help.
Just
you
know
to
to
get
everything
together.
D
And
that's
what
I
wanted
to
bring
up,
but
it's
a
perfect
lead-in.
I
was
gonna.
Do
you
guys
mind
if
I
work
on
that
high
level
diagram
and
at
least
get
us
started
on
that.
D
Okay,
sure,
all
right,
I
will
get
I'll
work
on
that
and
get
it
out
there
right
now.
This
dude
is
a
google
doc.
I
may
do
it.
I
may
just
skip
that
and
go
right
to
a
markdown,
so
we
can
have
it
as
part
of
our
documentation
piece.
D
So
I'll
figure
that
out
and
what
I
like
to
do
is
have
everybody
review
it.
So
we
can
make
sure
I'm
not
missing
anything.
D
Adam
takes
a
look
at
it
as
well,
adam
and
eukarys,
because
they're
going
to
be
working
on
the
the
captain's
side.
B
D
Okay,
good
yeah.
I
have
to
get
an
update
from
him
because
I
know
he
he
we
talked
to
last
week
last
friday
about
kind
of
the
overall
same
thing
we
were
talking
about
today.
You
know
how
the
process
is,
what
tools
we're
going
to
do.
D
What
so
I'll
put
that
together
in
a
diagram
and
kind
of
lay
out
one
of
the
use
cases
you
know
of
basically
what
we
talked
about
just
now
on
how
we
one
of
the
implementations
of
git,
ops,
now,
there's
gonna
be
different
ways
to
slice
it,
but
I
think
it'll
be
a
good
overall
view
of
all
three
tools,
working
together
or
four
tools.
If
you
can't
get
as
a
tool.
D
And
then-
and
I
think
I
owe
you
some
values-
files
brad
and
I
met-
I
will
try
to
get
those
early
next
week
taken
care
of.
So
you
have
some
values:
files
in
the
git
repos
that
are
real-
that
has
some
real
data
in
them
and
and
I'll
make
sure
that
they're
matching
up
with
what
is
in
the
templates
sure,
because
I
think
some
of
the
values
files
have
some
extra
junk
in
them
and
I'll
get
rid
of
all
that
stuff.
So
it's
easier
for
us
to
work
with.
C
Yep
sounds
good,
so
then
post
that
we
can
also
create
a
template
values
where
the
comments
are
so
you
will
always
have
to
have
one
template
value
style
where
the
comment
share
is
just
an
environment,
variable
or
dollar
value,
and
then
what
your
captain
will
do
is
always
go
and
see
that
and
replace
that
with
the
respective
environment
name,
so
we'll
keep
replacing
the
respective
file.
But
technically
it's
not
replacing
the
whole
file.
It's
just
replacing
one
line
right
exactly.
D
And
what
I'll
do
is
I'll
at
least
get
the
a
base
values
file
out
there
and
if
we
need
to
add,
like
you
said
variables
for
the
different
stages
of
the
pipeline,
we
can
do
that
down
the
road.
But
let
me
just
get
a
base
out
there
for
you
guys
to
start
working
with
sure.
B
D
B
B
Yeah,
it's
quite
scary
actually,
because
the
defaults
are
like,
sometimes
in
premium,
they
cost
like
400
bucks
a
month
for
the
this
application
gateway,
for
example,.
D
Yeah
and
I've
run
into
that
before,
like
with
our
our
database,
it
was
like
this
insane.
It
was
like
a
eight
processor
database.
When
you
know
all
we
need
is
a
two
core,
so
you
you'll
be
able
to
the
the
database.
Will
we
can
share
between
the
clusters
so
that
we
don't
have
to
scale
another
database,
we'll
be
good
with
sharing
that
across.
B
All
the
clusters-
that's
a
postgres
right,
yeah,
yeah,.
D
No,
it's
okay,
we'll
just
it'll
just
connect
up
to
the
same
schema.
D
No,
no,
the
way
it's
designed
it's
not
going
to
step
on
top
of
each
other.
We
can
have
multiple
clusters
pointing
at
the
same
database.
B
Perfect,
okay,
so
you
use
a
connection
string
or
identity
for
this
one.
D
B
Okay,
I
can,
I
can
go
and
hunt
for
them
and
then
just
replicate
them
into.
C
D
D
They're
gonna
be,
if
you
go
back
up
to
the
resource
group,
the
cluster
are
yeah.
I
think
they're
in
there
in
that
key
vault.
C
D
Yeah
it'd
be,
it
would
be,
I
think,
that's
great
to
know,
and
I
think
we
should
just
use
the
the
generic
out
of
the
box
helm
templating
for
now
and
then
once
we
get
everything
kind
of
put
put
together,
add
in
customized
I'm
very
curious
how
customized
is
gonna
work
with
argo.
C
I've
already
started
using
customize
with
argo.
Now
it's
pretty,
it's
works
pretty
fine.
The
only
problem
is
normally
helm
templates,
remove
a
lot
of
extra
work
which
you
might
have
to
do
within
our
go
or
your
help.
Your
customized
templates,
because
there's
a
lot
of
stuff
which
helm
just
takes
out,
takes
care
of
out
of
the
box
which
customize
wouldn't
then
you
would
have
to
overwrite
or
write
an
overlay,
etc.
C
B
They
are
I'll
filter
by
themselves.
D
Didn't
help
much
there's
like
the
the
pg
e
cred,
the
nginx
secret.
B
D
Yeah
and
when
it
when
the
helm,
charts
run
and
the
containers
start
up,
it
should
connect
and
we
should
be
up
and
going
on
those.
So
and
eventually,
what
we'll
do
is
we'll
have
to
figure
out
down
the
road.
If
we
want
to
stick
with
the
devops
cluster,
we
basically
down
the
road.
We
need
to
merge
the
cluster
sort
of
start
paying
for
the
extra
resources.
D
B
B
D
B
D
B
D
B
D
That's
what
I'm
kind
of
leaning
towards
once
we're
happy
with
with
this
this
one,
then
we'll
take
the
other
the
website
and
the
documentation
and
move
them
over
into
the
new.
The
devops
cluster.
B
Yeah
because
we
yeah
we
can
leverage
front
door
as
well,
where
we
need
to
yeah
I've
been
looking
all
of
my
security
or
all
the
security
and
stuff,
that's
good
for
the
other
one
as
well.
So
so
is
this
one.
D
Yeah,
I
totally
agree
the
the
more
we
can
slim
this
down.
I
think
the
better
off
we'll
be.
D
B
Okay,
cool.
D
Okay,
so
you
committed
to
that,
oh
go
to
you!.
D
D
Yeah
that'd
be
one
way
or
they're
in
in
the
organization.
There
is
teams.
B
D
B
What
I
might
adam,
I
tried,
I've
been
talking
to
adam
he's,
been
a
bit
quiet,
but
when
he
comes
back
online,
I'll
I'll,
add
him
as
well.
So
I
can
take
it
in
tickets,.
D
B
So
for
action
items
for
next
time
I
mean
we're.
Still
I
was
saying
I've
set
up
captain.
B
C
Once
we
set
up
the
poly
repo,
the
so
I
I
was
handling
the
argo
events,
part
right
so
will
will
the
captain
guy
handle
the
captain
part
like
yeah.
D
Yeah
so
ukrs
and
adam
will
handle
the
captain
events,
part
and
and
I'm
working.
C
E
D
I'm
working
with
both
of
them
to
do
the
integration
with
ortelius
as
well.
B
Very
we're
still
very
interested
in
the
events
as
well
yeah
for
that
person.
B
D
Yeah,
if
he
yeah,
he
has
to
accept
the
invite
and
then
it
should
show
up.
B
Okay,
can
you
accept
all
right,
we'll
just
make
sure
now
we
actually
have
to
finish
the
polling
reversal,
which
will
be
fine?
We
might
it
looks
like
it's
all
consistently.
You
know
in
the
same
chart
folders
with
the
values
that
young
on
everything,
so
we
shouldn't
have
any
problems
here.
B
So
we'll
we'll
get
what's
in
this
working,
we'll
I'll
make
one
ticket
to
to
change
over
the
secrets,
yeah
yeah
or
copy
something.
D
I'm
gonna
do
a.
Let
me
do
an
issue
for
creating
the
poc.
D
So
ujwal,
if
you
could,
in
the
ortillius
repo
or
two
repo,
if
you
could
just
update
the
readme
file
and
do
a
pr
against
that
and
when
you
do
that
it
allows
us
to
see
and
associate
your
user
id
to
the
organization.
D
A
I
think
tracy
we
were
talking.
She
wants
me
to
be
more
like
a
corporate
contributor.
D
D
D
E
C
D
Yeah
and
one
of
the
things
I
usually
do
like
on
our
our
build
side,
even
though
they
are
public
repos,
we
had
to
add
in
a
an
ssh
key
to
have
to
clone
them
without
any
weird
password,
props
and
stuff,
like
that,
so
you
may
run
into
where
you
need
an
ssh
key.
Just
let
me
know,
and
those
are
already
I
have
those
already
in
places.
So
it's
just
a
matter
of
getting
them
to
the
right
cluster
and
stuff.
C
D
C
Just
need
to
add
it
into
the
what
we
call
argo
secrets,
so
argo
directly
will
read
the
secrets
and
we
can
just
create
it
using
that
perfect.
C
So
that
way
you
can
directly
just
update
it.
Should
we
be
using
see
no
sealed
secrets
right,
just
a
normal
secret.
D
I've
been
just
using
opaque
secrets,
kubernetes.
B
D
That
exist
on
our
google
side,
so
there's
a
google
and
it
has
to
deal
with
some
user
ids
at
that
level,
but
during
our
so
we
use
google
cloud
build
as
the
the
build
process
and
there's
a
couple
logins
that
we
need
to
do
into
deploy
hub
and
artelius,
and
those
are
sealed
secrets
on
that
side.
But
those
are
coming
from
the
gcp
platform
for
the
azure
side.
We
should
just
do
for
everything
on
that.
We
should
just
do
opaque,
kubernetes
secrets.
C
Okay,
so
no
comments
into
the
thing,
because
there
are
two
ways
of
doing
it:
right:
brad
can
click
on
settings.
C
Repositories,
so
you
can
either
go
and
manually,
create
the
repository
here
and
you
put
in
the
ssh
key
or
the
other
option.
Is
you
basically
feed
it
into
the
github
repo
and
it
automatically
does
all
this
for
you.
So
that's
the
reason
I
was
asking
right
so
either
either
you
will
have
to
manually
create
that
respective
ssh
key
here,
so
you
go
and
create
a
new
password
structure
or
we
would
have
to
commit
that
same
secret
into
github
so
that
it
automatically
creates
all
of
this.
D
Okay,
let
me
just
double
check
the
ssh
key
which
one's
being
used
and
make
sure
it's
not
my
my
personal
private
key.
B
C
D
Okay,
that's
interesting.
I
I
have
not
played
around
with
the
pgp
keys
other
than
for
signing
my
commits.
That's
the
one.
The.
B
B
C
D
B
D
B
C
A
I
have
some
ammos
that
I
used
recently
for
istio
setup.
If
you
want,
I
can
share
them
with
you
like.
I
have
yeah
that'll,
be
super
helpful
yeah.
I
have
a
big
block,
so
so
one
of
the
things
with
istio
is
that
when
you
try
to
install
steel
it,
there
are
a
lot
of
add-ons
that
comes
with
the
stereo.
There
are
different
levels
of
setup.
You
can
do.
A
So
I,
but
I
was
installing
a
steel
on
a
cluster
where
I
already
had
bunch
of
stuff
previously
installed
like
prometheus,
and
you
know
prometheus
and
even
log
stash
and
things
like
that,
so
I've
done
like
a
base.
Just
like
a
a
regular
istio
install.
I
think
that
that
is
what
we
are
looking
for.
Right
yeah
are.
C
A
Can
I
share
my
screen
with
you
like?
Actually,
let
me
show
you
the
document.
B
I
just
put
the
issue
if
you
feel
free
to
add
some
stuff
in
there.
If
you
like.
A
So
I
I
did
it
specifically
on
aks
one
of
my
team
members,
so
istio
ctl
installation,
and
then
then
there
is
an
sdo
operator.
You
need
to
install,
so
you
can
just
follow
this
document.
So,
like
I
said
right
there,
there
are
profiles
within
hto.
A
You
can
set
up
like
within
these
profiles,
like
you
can
enable
add-on
components
like
grafana
prometheus
other
things,
but
I
already
had
this
setup
previously,
so
I
had
to
like
comment
out
all
of
this
and
yeah,
so
I
can
clean
up
this
document
a
little
bit
and
send
it
over,
and
you
just
follow
this
document
and
you
should
be
able
to
install
istio
and
kiyali
as
well.
Keal
is
primarily
for
tracing
if
you
want
to
see
tracing
between
microservices.
A
So
for
aks
I
had
a
couple
of
issues
with
the
so
I'm
using
a
steel
cli,
istio
ctl,
which
is
cli
to
initiate
the
operator.
A
B
C
Yeah
sure
I
will
send
you
my
calendly
and
then
that
way
we
can
just
keep
in
touch,
and
you
know
if
we
need
to
book
each
other's
time.
It'll
be
easy,
okay
or
yeah.
I
I'm
also
on
I'm
supposed
to
join
the
discord
channel
for
your
switch-ups
are
not
done
so.
I
need
to
do
that.
B
B
B
D
And
we're
making
some
good
progress.
I
like
it.