►
From YouTube: April 29th Architecture Meeting
Description
Updates from the Service Catalog project and GitOps Project is included in this April 29th Ortelius Architecture meeting.
A
A
If
that
time
works
out
better
for
you,
I
think
it
works
out
to
be
like
eight
or
eight.
I
think
8
30,
australian,
probably
sydney
time
zone.
So
that's
just
a
heads
up.
So
just
a
quick
couple,
things
sergio.
I
think
your
your
your
little
tutorial
how
to
stand
up
the
the
mini
cube
with
ortulius
in
it
was
was
brilliant.
B
A
So
you'll
have
to
you'll
have
to
give
us
all
the
a
list
of
all
the
tools
that
you
used
and
where
the
resources
were
so
I'll,
open,
a
github
or
there's
a
issue
out
there
I'll
go
ahead
and
update
it
and
mention
you
in
it.
So
we
can
get
a
list
of
tools
that
you
used
on
the
architecture
front
had
a
couple
of
interesting
conversations
this
week
with
some
folks
and
one
of
the
things
that
has
come
up.
A
You
know
like
tracy,
goes
and
does
these
interviews
with
people
to
find
out
what
they're
doing
around
microservices,
and
I
can't
remember
if
it's
a
devsec
ops
person
or
if,
if
it
was
like
a
developer
role,
but
it
what
came
up
was
swagger
and
the
restful
api
sets,
and
you
know
we're
building
out
the
service
catalog
with
the
things
like
the
pagerduty,
the
slack
channel
for
the
developers.
You
know
what
what
is
the
slack
channel?
A
C
There's
a
there's
a
load
of
options
around
it,
and
so
we
use
swagger
quite
a
lot
internally,
but
to
be
honestly
even
even
doc,
see
if
I
remember
I
can
render
it
so.
Swagger
and
open
api
are
generally
pretty
interchangeable.
Nowadays.
The
tools
which
work
with
one
worked
for
the
other
swagger
kind
of
evolved
into
open
api.
D
C
Dig
out
some
of
the
stuff
we're
using,
but
yeah
you
can
render
it
online.
You
can
also,
generally
most
libraries
will
support
standing
up
essentially
a
swagger
endpoint,
alongside
with
your
service,
which
exposes
the
api
on
a
specified
port
and
that
swagger
is
really
just
static
content.
At
that
point,.
A
C
So,
probably
that's
not
from
there
so
swagger
when
you
end
it,
it
does
also
help
a
json
file
which
might
be
easier
for
us
to
consume
and
use
something
I
think
from
democracy
spotify's
backstage
which
is
and
when
I
say,
react
based
and
they
have
a
part
of
their
technical
documentation
module
for
that
points.
A
you
point
it
to
get
repo
with
a
swag
location
in
it
and
it
can
go
and
it
can
link
to
that
and
render
some
of
that.
A
E
E
E
C
Just
just
to
expand
that
there's
also
slightly
different
way
of
interacting
with
swagger,
which
kind
of
makes
this
kind
of
discovery
even
more
powerful
or
more
useful
or
more
it's
a
missing
gap.
So
the
way
we
use
swagger
is
we
don't
write
the
swagger
yaml
file,
at
least
in
the
dot-net
code
or
the
python
code.
There
is
a
swagger
library
of
the
open
api
library
we're
using
allows
us
to
decorate
the
methods
we're
defining
and
it
generates
for
swagger
off
the
back
of
that.
C
So
it's
not
a
case
of
writing
yemen
initially,
then
implementing
it.
It's
actually
decorated
in
with
the
code,
the
tricky
bit
with
that
from
a
kind
of
ongoing
life
cycle
perspective
is
discovering
swagger
discovering
the
api
endpoints.
If
someone,
if
it's
not
being
published
to
somewhere,
which
is
easily
discoverable
teams,
add
new
features
to
the
service.
C
Sometimes,
contract
testing
isn't
in
place,
but
it
is
in
the
swagger
because
it's
being
generated
automatically
whenever
there
is
a
build
of
that
code
or
a
run
of
that
code.
So
we've
run
into
situations
where
the
lack
of
a
central
repository
until
recently
has
been
a
major
issue
with
consumption
of
services
when
someone's
added
something
and
then
it
hasn't
been
well
shared.
A
A
If
we
could,
in
the
scenario
that
you're
talking
about
owen,
bring
in
the
swagger
yama
or
the
swagger
json,
whichever
one
works
out
better
and
bring
that
in
and
persist
persisted
in
our
database
alongside
the
component
version,
which
has
the
link
to
you,
know
the
get
repo
the
the
docker
registry
kind
of
centralizing
everything
and
then
chrome
in
your
case,
where
you
have
like
swagger
hub.
That
would
just
be
where
we'd
be
pointing
to
the
url
for
that
version
of
the
of
the
api
in
the
swagger
hub.
F
Does
anybody
had
experience
actually
the
normalization
of
the
swagger?
I
mean
what
if
two
teams
come
up
with
actual
same
api
by
maybe
they
will
end
up
with
a
different
swagger
yammer
file.
This
is
possible,
or
is
it
possible
to
normalize
it
simply
find
out?
This
is
unique
definition
of
service
just
to
make
sure
that
this
is
actually
what
we
want
to
really
index
right.
You
want
to
make
very
so,
let's.
F
A
Well,
that
would
be
that
would
be
an
interesting,
downstream
add-on
to
the
what
we're
talking
about
would
be.
You
have
the
rest.
You
know
the
the
restful
api
endpoints
that
this
microservices
is
publishing
and
then,
like
you,
you
said
taurus.
The
other
side
is
what
are
the
restful
apis,
that
this
microservice
is
consuming?
B
B
A
So,
oh
and
I'm
going
to
kind
of
throw
a
little
bit
research
over
your
way
to
see
what
tools
we
should
be
looking
at.
You
know
if
this
is
going
to
be
a
node.js,
if
you
want
to
how
we
want
to
interact
with
it,
since
he
has
some
experience
on
that
side
and
then
chrome,
if
you
can
kind
of
lay
out
how
you're
using
swagger
hub,
because
I'm
totally
unfamiliar
with
swagger
hub
and
I'd
like
to
find
out
how
that
would
we'd
handle
that
scenario
as
well.
C
Sounds
good
to
me,
sorry,
there's
something
which
I
know
linkedin
does,
but
I
don't
know
if
istio
does
need
to
look
into
it.
So
linkedin's
got
a
really
interesting
option
where
it
can
actually
import
a
swagger
and
open
api
spec
and
use
that
to
enrich
the
service
catalog.
I
don't
know
if
seo
has
a
similar
functionality.
I
expected
something
along
those
lines,
but
it
might
be
worth
being
aware
of
when
we're
looking
at
the
service
catalog
in
case
seo
does
have
that
element
to
it,
too.
C
C
A
B
A
D
C
D
D
It's
not
a
huge,
it's
not
a
huge
technical
effort.
I
was
thinking
that
maybe
we
could
get
you
know.
Powell
wants
to
do
more
coding
and
that
might
be
a
good
place
for
him
to
start
to
start
getting
an
understanding
of
how
things
are
fitting
together.
But
steve.
Could
you
talk
about
what
we
need
to
do
to
get
that
done.
A
Yeah,
so
the
code
is
currently
commented
with
the
a
feral
2
license
and
it
has
a
copyright,
I
think,
to
one
of
our
companies
called
catalyst
systems
corporation
or
open
mic
software.
I
can't
remember
how
it's
copyrighted,
but
we
need
to
change
that
to
an
apache
2
license
so
any
of
the
source
code,
which
has
the
header
at
the
top.
A
D
A
I
have
tried
in
the
past
to
run
tools
that
little
programs
go
strip
and
replace
copyrights,
but
they
don't
always
necessarily
work
very
pretty
so
it's
gonna
be
a
little
bit
of
poking
around
and
doing
a
cut
and
paste
along
those
lines.
So
that's
kind
of
what's
happening
on
that
front.
The
at
the
repo
level
you
know
flipping
the
license
is
just
a
matter
of
doing
a
pr
at
that
level.
So
that's
not
going
to
be
an
issue
there.
A
The
only
problem
that
we'll
have
is,
if
the
license
change
takes
too
long
to
do
we'll
get
it.
We
we
have
the
potential
to
get
some
merge
conflicts
that
we'll
need
to
resolve
it's
manageable,
but
that's
a
possibility
that
we
can
run
into
so
that's
kind
of
what
we're
looking
at.
If
anybody
knows
of
those
little
tools
to
do
the
search
and
replace
that
actually
work,
if
you
can
throw
them
my
way,
that'd
be
great.
C
A
A
A
What
we
have
is
at
the
very
top
is
the
license
here,
so
this
section
and
every
single
file
for
all
all
the
java
code
needs
to
be
changed.
This
is
referring
to
the
ferrule
license.
It
would
have
to
be
changed
over
to
the
the
gnu
license
and,
while
we're
in
here,
we
should
probably
do
the
spdx
tags
as
well
to
be
more
compliant
with
the
the
packaging
on
that
front.
A
So
putting
in
the
replacing
the
the
the
verbiage
with
the
apache
2,
attributing
it
to
the
linux
foundation
and
putting
in
the
spdx
tags
is
something
that
we
would
do
now.
This
would
happen
on
all
the
different
files,
so
like
login,
will
have
something
very
similar,
they're,
pretty
well
standardized
like
in
the
java
code.
But
if
you
get
into
some
of
like
the
c
code,
it's
gonna
be
a
different
world.
A
A
A
A
Yeah,
I
think
it's
exactly.
I
think
it's
gonna
be
a
mix
of
tools
and
I
think
also
it's
just
gonna-
be
a
cut
and
paste
effort
on
some
of
it,
and
I
think
we
can
get
some
people
that
are
new
to
the
coding
world
to
help
us
out
on
this
kind
of
divide
and
conquer
the
process.
So
we
can
have
somebody
do
the
c
code.
We
have
somebody
to
do
the
java
code,
somebody
to
do
the
python
kind
of
split
it
up
that
way.