►
From YouTube: Ortelius General Community Meeting - August 11, 2020
Description
In this meeting we discussed project goals, working groups and what is the next steps for building the Ortelius open source project for mapping microservices and building a shared catalog of deployable services.
A
A
We
know
that,
as
we
start
moving
down
this
microservices
architecture
that
we're
going
to
be
sharing
them
in
the
same
way
as
we
share
open
source,
think
about
an
api
marketplace,
we'll
have
a
microservice
marketplace
and
then
the
final
one
is
to
really
focus
on
providing
visibility
and
reporting
into.
What's
has
what
has
been
deployed
across
clusters.
A
A
lot
of
solutions
right
now
are
creating
observability
reporting
around
a
single
cluster,
but
what
we
want
to
do
is
set
it
above
the
cluster
level
and
really
look
to
see
what's
been
deployed
and
the
configuration
of
the
microservices
at
a
particular
cluster
level,
there's
some
working
groups
that
we've
defined,
including
the
what
we're
calling
the
cd
environment.
This
is
probably
our
most
important
one
that
we
need
to
get
working
on
first
and
that
cd
environment
really
is
is
our
own
development
infrastructure
that
will
help
us
be
more
productive.
As
a
group.
A
A
second
working
group
is
data
science
and
visualization.
We
desperately
need
that
for
goal.
Number
four,
which
is
really
creating
those
clean
reports
and
maps
deployment
integrations,
relates
back
to
what
we
need
to
do
for
supporting
these
open
source,
cd
tools
and
I've
added
argo
to
that.
Even
though
it's
in
the
cncf
we're
okay
with
that,
and
then
there
is
creating
that
marcus
marketplace,
the
marketplace
is
going
to
be
the
the
the
piece
that
really
drives
this
project.
A
I
have
a
strong
feeling
that
being
able
to
share
and
and
define
microservices
star
them
create
risk
analysis
around
them
will
be
a
really
critical
part
of
this
project.
We
need
a
team,
that's
willing
to
take
on
usability
and
testing,
we
need
a
team
for
doing
documentation
and
we
need
a
team
for
architecture.
A
Architecture
will
be
extremely
important
as
we
move
forward
and
start
thinking
about
how
to
actually
break
ortilius
down
into
to
into
individual
microservices,
and
then
we
just
need
a
good
development
team
who's
willing
to
take
on
what
we
need
and,
of
course,
there's
always
outreach.
We
need
to
do
some
work
on
our
logo
on
our
website
and
we're
thinking
about
an
alien
for
a
logo
and
then
project
management.
A
A
B
B
So
what
ends
up
happening?
Is
we
end
up
publishing
a
command
line
program
to
to
pi
pi?
That
allows
us
to
have
a
python
module
the
command
line,
but
we
go
and
and
wrapper
that
into
a
container
just
because
what
we
found
is
a
lot
of
people.
Don't
have
python
installed
in
their
build
process,
but
they
do
have
docker
so
we'll
wrapper
that
into
a
container
and
allow
that
to
be
utilized
now.
Things
like
jenkins,
x,
tecton
and
I
believe
argo
are
all
want
to
even
like
google
cloud
build.
B
They
all
want
to
run
containers
as
part
of
the
cd
process,
so
this
one
is
where
we're
going
to
be
extending
out
that
model
and
hooking
into
all
the
different
tools
there.
So
if
you're
familiar
with
any
of
these,
please
sign
up
like
I,
I
know
what
spinnaker
is,
but
I've
never
really
played
with
it.
Argo's
relatively
new.
It
got
argo,
got
placed
into
the
cncf
side
of
things,
but
from
my
understand,
it's
a
pretty
kubernetes
heavy
type
of
solution
and.
A
B
A
B
Yeah
and
the
kind
of
the
goal
would
be
to
have
a
single
code
base
that
we
can
just
tweak.
You
know,
maybe
it's
a
different
docker
build
file.
You
know,
but
basically,
underneath
the
the
covers.
It's
the
same
code
base,
doing
all
the
integrations,
and
then
we
just
have
you
know
like
technique
have
to
create
a
cdr.
Jacob's
x
is
the
same
way
and
there
are
slightly
different
cdrs
depending
upon
that,
but
we
can
create
different
containers
underlying
it's
going
to
be
the
same
code
base.
B
That's
doing
all
the
heavy
lifting
the
marketplace
and
domains
like
I
said,
I'm
actually
on
the
the
tecton
catalog
and
hub
sub
project
kind
of
been
peeking
in
on
how
they've
been
doing
that.
One
of
the
challenges
that
I
foresee
in
the
microservice
world
is
the
ability
to
really
structure
these
in
a
domain
driven
design.
B
B
So
and
it's
just
moving
in,
they
just
moved
it
out
experimental
and
they
have
several
level
different
levels
of
tiers
and
that's
what
they're
they're
working
right
now
they
have
like
a
a
community
tier
then
they
have.
The
next
tier
is
going
to
be
a
a
base
certification
and
then
they
have
like
a
full
certification
and
maintained
by
techton
for
the
different
tasks.
B
So
a
task
is
just
going
to
be
a
step
in
in
the
the
work
in
the
pipeline
or
the
workflow,
so
they're
trying
to
do
that
organization
as
part
of
that
right
now.
They're
they're,
currently
thinking
that
they're
just
gonna
have
one
catalog
for
everybody,
but
I
could
already
see
the
writing
on
the
wall
that
people
are
going
to
want
to
have
their
own
private
catalogs
of
tasks
that
they
want
to
run
internally
as
part
of
that
process.
So
that's
one
of
the
things
on
the
marketplace:
domain-driven
design.
B
B
The
ux
and
this
kind
of
comes
into
the
ux
and
testing
are
kind
of
hand
in
hand.
Here.
How
do
we
make
the
the
whole
ortelius
easier
to
use?
You
know
how
do
we
get
somebody
up
and
going?
You
know
how
we
can
make
the
navigation
easy.
Are
we
reporting
and
visualizing
the
the
data
that
we
have
easy
for
for
people
to
use
the?
B
So
I
think
we'll
throw
in
the
the
testing
in
that,
because
the
testing
we
have
the
two
parts
of
the
testing
one
will
be
more
of
the
well
three
parts.
We
have
the
cli
type
of
testing,
we
have
the
api
testing
and
then
we
have
the
ui
testing
process,
because
what
I'm
kind
of
foreseeing
is
that
when
we
start
throwing
some
test
cases
at
the
ux,
that
will
will
recognize
that
there's
certain
patterns
that
we
need
to
go
through
and
redesign
the
ux
to
make
things
easier,
documentation,
there's
a
documentation.
B
B
C
B
C
B
B
A
More
than
likely,
we
will
probably
keep
it
on
google,
and
I
say
that
because
if
we
are
moved
into
a
sandbox,
the
cdf
we'll
then
need
to
start
covering
the
cost
of
that
which
would
be
hugely
helpful.
And
right
now,
google
has
been
putting
a
lot
of
they've
been
hosting
a
lot
of
those
types
of
projects
on
behalf
of
the
cdf
for
free.
B
So,
and
it
may
be
a
split
where
we
actually
are
doing
the
pipeline
on
on
google,
like
you
said,
maybe
with
github
actions
and
mixing
the
two
together
but
deploying
out
to
azure
as
the
area
to
run
the
test
cases.
You
know
stand
up
a
couple
clusters
over
in
azure
for
different
test
cases
that
we
would
need
to
run
as
part
part
of
that
process.
B
B
B
And
one
of
the
things
that
we'll
probably
do
is
one
of
the
first
things
I
think.
There's
like
one
or
two
micro
services
that
come
off
the
top
of
my
head,
that
we
can
actually
start
with
that
are
kind
of
independent
from
the
rest
of
the
code
base,
and
I
was
thinking
as
just
like
a
little
exercise
is
to
like
one
of
them
is
to
display
the
log.
B
B
So
one
would
go
get
the
data,
the
other
one
go
ahead
and
render
it
to
the
browser.
But
we
could
do
like
rust.
Python
node.js
go
and
do
those
like
as
a
little
competition
to
see
which
one
which
microservice
language
seems
to
fit
best
and
the
easiest
to
use
as
part
of
that
creating
a
microservice
process.
B
So
that
was
one
thing
I
was
I
was
thinking
about,
was
throwing
that
out
there
and
having
different
people
that
are
familiar
with
those
languages
really
set
that
up
and
kind
of
review
for
everybody
else.
You
know
what
it
took
to
to
make
that
microservice
actually
happen,
and
if
there's
any
other
languages
I
missed,
I
mean
we
could
go
back
to
see
an
assembler.
I
know
those,
so
we
can
actually
go
back
to
that
world
as
well
as
some
fun
stuff,
then
just
general
development
development
work.
B
You
know
if
you
guys
have
enhancements
that
you
see
that
are
needed
or
bug
fixes.
Let's
get
those
up
to
date
out
on
the
issues
right
now,
just
hang
them
off
of
the
core
ortelius
repository,
instead
of
unless
it's
specifically
for
a
dock.
If
it's
a
dock
change,
let's
go
ahead
and
put
tie
it
to
the
dock
repo.
But
if
it's
like
a
bug,
fix
or
enhancement,
let's
go
ahead
and
start
it
in
the
in
the
core
repo,
so
christopher,
your
little
enhancement
to
pull.
You
know,
parse
the
the
aws
information.
B
I
say
we
go
ahead
and
put
that
as
an
enhancement
out
in
the
the
core
repository
and
then
we
could
track
that
and
get
the
coding
changes
in
and
that
may
start
out
in
the
core
repository,
but
maybe
get
maybe
moved
over
to
the
the
command
line
interface
world.
But
we'll
start
in
the
core
repository
with
that
that
type
of
issue
or
enhancement
the
website,
branding
messaging
right.
Now
we
have
the
little.
I
can't
remember
what
it's
called,
but
the.
B
Yeah
you
know
ortelius
abraham
martilius
was
the
first
person
that
created
the
world
atlas,
so
that's
kind
of
the
thought
process
because
it
ties
into
the
mapping
that
we
have
around
the
relationships
and
microservices
along
with
that
in
the
in
that
whole
world,
there
is
really
pushing
out.
You
know
letting
people
know
that
we're
actually
what
we're
up
to
and
the
more
the
more
we
more
visibility
that
we
have
the
more
chance
that
we
have
to
get
into
the
cdf
easier
things
more
resources
we'll
get.
B
B
I
actually
found
a
couple
bugs
literally
bugs
one
was
a
praying
mantis
and
also
I
had
a
a
baby
tarantula
in
my
office
and
you'll,
see
on
the
discord
channel
where
we
had
the
jenkins
sticker,
and
that
was
a
logo
that
they
did
around
an
alien
as
part
of
that.
So
maybe
we
could
have
an
alien
mapping
out
the
world
or
the
galaxy,
so
just
some
thoughts
and
then
just
on
the
project
management
side.
A
A
A
B
A
A
B
Right
now
we
have
that,
like
I
said,
we
have
not
set
up
the
a
public
facing
or
a
community-facing
work
environment.
Yet
right
now,
it's
all
internal
to
deploy
hub
and
that's
one
of
the
things
that
I'm
in
the
process
of
work
of
moving
over.
So
there's
things
like
that.
We're
working
on
like
getting
a
zoom
account.
A
C
A
C
B
B
Out
there
is
the
artelius
ortelius
is
the
is
basically
the
source
code
repository,
and
this
is
where
we
have
the
issues
and
pull
requests
that
we're
currently
working
on.
This
is
what
I
say,
and
we
just
have
the
the
few
contributors
right
now,
if
you
can
add
yourself
as
a
contributor
to
this
one,
then
you'll
at
least
get
noticed.
If
we'll
funnel
all
the
the
issues
right
now
will
funnel
through
this
one,
and
then
we
may
move
them
into
the
other
repositories.
B
B
A
B
And
then,
if
it's
a
doc
thing
you'll
see
out
here
the
docs-
and
this
is
all
the
markdown
and
how
to
run
the
markdown,
so
the
hugo
server.
Let
me
know
if
you're
running
on
windows,
the
installing
hugo,
you
need
to
install
the
extended
package.
Let
me
know
if
you
run
into
any
issues
getting
hugo
installed
for
running.
B
A
B
But
that's
where
the
docs
are.
You
know
we
have
a
couple
things
for
documentation
changes.
You
know
simple
things
like
markdown
needs
to
be
cleaned
up.
We
had
this
one
come
in
and
we're
where
to
start
type
of
where
to
start
with
ortelius
and
then
the
other
pieces
that
we
have
out
here.
The
comp
update
is
a
cli
interface.
B
All
these
one
with
stores
this
is
the
hipster
store
repository.
So
this
is
the
sample
code.
This
goes
along
with
the
sample
code
and
then
here's
some
of
the
other
repositories.
This
is
for
when
we
created
the
cdr
for
nebula,
which
is
now
called
relay
as
part
of
the
puppet
world.
I
have
to
see
with
what
they
I
have
to
circle
back
around
with
puppet
to
see
what
they
have
going
on
on
that
front
and
then
just
some
more
demo
stuff.
B
Things
that
I
know
we
need
to
do
is
add
in
owner's
files,
so
you
guys
can
be
reviewers
and
all
that
stuff,
so
just
bear
with
us
as
we
get
that
sorted
out
and
if
you
want
to
be
added
to
any
of
this
as
a
reviewer
or
you
know
the
different
levels,
just
let
me
know
and
we'll
get
that
taken
care
of
we're
early
on
in
this
process.
So
we
have
the
opportunity
to
really
form
this.
The
way
it
makes
sense
for
you
guys.
E
Okay
yeah,
so
I
was
just
asking
that
you
said
that
we
are
going
to
run
the
test
infra
on
azure,
so
but
on
the
other
hand,
we
you,
you
said
that
the
our
our
python
code
is
being
packaged
is
as
docker
and
it's
going
to
be
run
on
kubernetes.
So
it's
is
that
the
exact
same
code
that
we
are
going
to
test.
If
yes,
then,
should
we
are
we
going
to
you?
What
kind
of
is
your
infra?
Are
we
planning
to
use
to
test
our
services?
E
B
B
You
know,
however,
we
want
to
set
it
up
and
if
we
need
to,
if
we
need
to
stand
up
like
a
a
vm,
a
vm
image
type
of
thing
that
we
want
to
run
the
test
cases
against
like
from
we
want,
like
a
windows,
desktop
that
we're
going
to
run
selenium
from
the
test
cases.
We
could
do
that.
You
know
it
doesn't
have
to
be
a
kubernetes
style
test
workload
whatever
makes
sense.
Let's
take
that
path,
that's
the
easiest!
B
A
D
A
A
So
what
I
am
thinking
is
that
it
would
be
right
now
if
we
look
at
what
is
the
most
important
to
get
started,
it's
this
top
one.
You
know
these
are.
These
are
will
come
in
time
and
if
anybody
wants
to
take
on
any
of
any
of
these
it's
great,
but
we
need
to
get
this
environment
sorted
out
for
the
community
so
that
everybody
can
have
a
consistent
way
of
working
and
testing.
B
A
F
I
I
just
have
a
question.
Sorry,
if
it's
irrelevant
on
this,
when
we
are
talking
about
testing,
are
we
talking
about,
like
all
the
services
are
already
deployed,
and
we
are
make
sure
that
they
are
tests.
They
are
all
running
healthy,
integrated
together
and
ui
can
talk
to
all
the
services
and
existing
micro
services,
or
we
are
talking
about
like
templatizing
or
micro
services
that
it's
already
taken
care
of
like
in
the
kubernetes
world.
F
What
is
like,
because
sometimes
we
don't
want
to
even
promote
it
to
the
next
level
until
I
I
my
component
is
fully
working
and
then
after
that,
test
is
integrated
well
and
then,
after
that,
ui
can
just
integrate
well
with
all
of
the
underlying
micro.
Is
this
considered
in
this
pipeline
all
of
them,
or
it
has
nothing
to
do
with
this?
It's
already.
We
already
talked
about
it,
because
this
is
my
first
time
I
don't
know.
A
A
C
A
C
Sides,
since
I
have
worked
on
means,
I
am
working
on
qa
testing
and
automation,
yeah,
so
the
I
guess,
the
first
level
checks
will
be
for
code
two-way
unit
test.
Right
means
our
developers
should
have
a
unit
test
repo,
it's
that
should
be
passed
on
every
pipeline
right
whenever
each
commit
goes,
that
test
should
get
passed,
and
after
that
we
should
have
a
infrastructure
level
of
code
as
well.
Right
suppose
the
version
is
mismatched
kind
of
thing,
so
that
categories
fall
in
the
integration
kind
of
test.
C
F
What
should
fail
on
this
service
fast,
because
sometimes
your
your
unit
test
and
component
test
is
keep
failing
in
the
environment
and
you
cannot
promote
it
and
then,
when
it's
just
start
overing
or
is
cascading
to
the
underlying
service
and
bring
down
the
other
service
event.
So
it
has
to
be
defined
in
the
whole
process
that
we
are
not
causing
every
service
fail.
Every
service
failed
because
the
dependency
it
cannot
talk
to
the
other
dependent
right.
E
A
Point
where
it
is
in
microservices,
however,
to
that
point
and
when
it
comes
to
testing
how
ortillius
is
working,
that
is
part
of
the
solution
that
ortilius
is
providing
and
that
that
actually
needs
to
be
clearly
tested
out
to
make
sure
that
the
version
tracking
is
is
at
is
accurate.
So,
for
example,
I
think
I
I
may
have
opened
a
ticket,
but
I
saw
a
place
where,
where
a
we,
I
had,
we
had
a
microservice
that
was
being
consumed
by
three
different
applications.
A
D
Tracy,
this
is
eric
just
into
anon's
point.
It
sounds
like
ortillius
is
going
to
evolve
into
many
types
of
different
test
case
scenarios
with
the
different
types
of
code
base
we
use,
might
have
different
unit
testing
frameworks
and
then
with
selenium,
that's
a
whole
other
can
of
worms.
You
know
is
that
going
to
be
done
with
j
unit
or
what?
What
platform
is
that
going
to
be
right?
So
I
think
it
sounds
like
it's
going
to
evolve
into
many
different
streams
of
different
types
that
I
think.
C
B
And
then
also
one
of
the
things
I
kind
of
envision
is
putting
a
istio
service
mesh
on
there,
so
we
can
actually
send
out
a
as
we
start
breaking
artelias
into
smaller
microservices
that
we
can
send
out
a
change
and
put
it
out
into
the
the
same
cluster
and
route
over
to
some
test
cases.
To
that
as
like
a
beta
or
whatever.
We
want
to
call.
F
B
F
B
You
know
when
we
start
looking
at.
I
have
a
service
change
and
it's
a
simple
update.
What
path
do
we
need
to
go
down?
You
know
and
that's
where
the
whole
ml
process
that
we're
talking
about
you
know
decision
making
as
part
of
this,
how
it's
going
to
affect
the
transition
from
a
developer's
change
out
to
a
production
run
time
you
know
which
which
path
we're
going
to
take,
because
not
every
not
every
change
needs
to
go
down
the
same
path.
All
the
time.
A
B
A
A
If
you
you
know,
if,
if
there's
an
area
that
you
feel
like,
you
would
like
to
put
your
name
to.
Let's
get
that
so
we
know
who
the
some
of
the
initial
team
is
going
to
be.
I
have
recorded
this
I'm
going
to
send
it
out
to
the
folks
who
could
not
come
I'm
going
to
ask
them
to
do
the
same
thing,
but
in
terms
of
what
we
really
need
to
get
done
and
from
just
a
nuts
and
bolts
level.
It's
this.
A
So
if
there's
anybody
else
on
this
call
today,
who
wants
to
help
with
getting
some
of
this
stuff
set
up,
let
me
know
or
add
your
name
there,
and
that
working
group
group
will
get
started
getting
this
sorted
out
so
that
the
rest
of
us
can
actually
start
playing
and
and
making
contributions
and
getting
the
getting
it
built
out.
All
of
that
fun
stuff.
C
B
D
A
A
I'm
gonna
I
have
just
I
have
to.
I
need
to
fix
this
one,
because
I've
already
started
uploading
to
it.
I'm
assuming
I'll
get
it
I'll,
get
it
sorted
out.
I
have
the
thing
I
just
have
to
be
able
to
input
it,
but
they're
like
it
took
too
long
it's
like
well.
I
don't
can't
help
that,
but
it's
listed
under
me
right
now
and
I
need
to
get
it
changed
to.
I
have
to
get
it
fixed,
so
that's
where
it's
at.
A
I
had
it
as
ortillius
and
they
for
some
reason
they
didn't
like
it,
and
the
only
way
for
me
to
keep
it
up
was
to
change
it.
To
my
account,
I
will
work
on
that
today,
I'll
get
it
fixed
somehow,
but
ortillius
itself.
I
think
it's
gone
well.
I
guess
not
yeah,
because
this
was
was.
This
just
is
that
was
that
just
a
I
guess,
that's
just
one
video
I'll
get
it
I'll.
I
will
sort
it
out
today.
A
D
A
B
A
A
D
E
B
It's
tied
to
your
organization,
so
techtons
run
into
this
problem,
because
google
originally
started
techton
and
they
had
the
google
meet.
So
it
has
to
be
a
google
employee
that
starts
the
meeting
and
recording
it.
Otherwise,
nobody
else
can
you
can
get
it
going.
Okay,
it
was
it's
a
weird
thing
that
we've
run
into
with.
They
don't
make
it
easy.