►
From YouTube: February 18, 2021 Ortelius Architecture Working Group
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
So
today
is
the
february
18th
architecture
meeting
and
go
ahead
and
sign
in
I
threw
the
link
in
the
chat,
welcome,
jim
and
zach.
I
know
there's
a
couple,
others
like
sasha
and
owen.
That
said
they
couldn't
make
it
today,
so
they
reviewed
the
doc
earlier
and
we'll
just
go
over
this
real
quick.
So
I
we've
been
some
of
the
things
that
we
have
been
working
on
last
last
meeting
was
around
application
sets
and
what
those
mean
to
us
and
how
that
would
fit
into
our
world
of
ortelius.
B
So
today
we're
going
to
look
at
service,
catalog
information,
and
let
me
pull
that
back
up.
So
when
we
talk
about
this,
the
service
catalog,
what
we're
seeing
from
interviewing
customers
and
people
and
looking
at
other
open
source
tools
that
are
out
there,
people
are
looking
to
gather
information
about
a
particular
microservice
things
that
we
found.
I'm
not
gonna
go
into
real
detail
on
these,
but
that,
like
the
build
information,
the
docker
information,
what
was
the
commit
and
then
like
who's,
the
owner
of
the
project?
B
If
this
service
breaks,
how
do
we
contact
that
owner
as
part
of
that
process?
So
what
is
really
looking
at
it
is
from
the
sre
operations
side.
Something's
broken.
I
need
to
get
a
get
in
touch
with
somebody
to
go
fix
it.
What
the
development
slack
channel
is,
or
the
discord
hip
chat
that
they're
using
as
that
process
and
then
information
about
the
pager
duty,
whether
it's
a
the
pager
duty
service,
the
the
flow
for
the
the
resolution
tree
at
the
service
level
or
at
the
business
service
level.
B
B
One
of
the
tricky
things
about
this
is
a
server.
A
microservice
can
be
deployed
to
multiple
environments,
so
we
could
have
dev
test
prod.
We
have
that
service
living
in
those
three
environments
and
then
within
each
one
of
those
environments,
we're
going
to
have
a
number
of
clusters
that
they're
going
to
be
deployed
to
and
then
within
that
cluster
there's
going
to
be
a
log
url.
B
So
it's
this
one
to
end
and
relationship
that
we
need
to
keep
track
of
and
one
of
the
things
on
that
is
being
able
to
really
point
out
from
an
sre
point
of
view,
here's
the
log
files
for
this
service.
We
know
that
this
service
is
having
issues
where.
Where
can
we
look
at
the
logs
as
part
of
that
so
like
in
the
the
google
world,
the
gke
world?
You
know
google
uses
stackdriver
as
pro
part
of
that
if
we're
looking
at
kubernetes,
maybe
a
prometheus
log
for
the
cluster.
B
So
that's
kind
of
the
thought
around
runtime
log
urls,
and
I
was
wondering
if
that
makes
sense.
If
anybody
has
any
comments
on
that
or
if
we're
going
down
like
the
totally
wrong
tree
path
on
this
one.
C
No
comments
pretty
reasonable
to
me.
You
know:
you've
got
the
traditional
logs
which
aren't
going
to
have
any
sort
of
sensible
url
associated
with
them,
but
those
aren't
very
meaningful
in
a
long-term
sense.
So
you
really
want
something.
That's
got
a
url,
a
bunch
of
that
stuff
lands
in
s3,
but
you
can
put
urls
on
s3
stuff.
So
all
that
seems
pretty
happy.
Okay,
cool.
B
B
You
know
the
reason
why
the
service
is
having
problems
is
there's
a
nginx
low,
balancer
problem,
but
you
know
that's
that's
causing
a
downstream
problem
or
something
like
that,
but
I
I
think
in
general,
if
we
could
just
give
some
information
to
make
it
a
little
bit
easier
around
that
was.
This
is
my
goal
on
that.
One.
D
I
got
a
curious
question
to
add
in
there
yeah
this
is
aaron
brooker's.
My
first
time
tracy
and
I've
talked
a
few
times,
so
I
always
think
about
it
from
a
perspective
of
multiple
hot
hot
deployments,
with
automated
failover
through
something
like
istio
or
console
some
of
the
observability
data
right
from
a
transactional
standpoint
of
which,
where
it
went
to
right
because
istio
will
have
that
automated
failover.
So
maybe
one.
D
B
Yeah
would
that
would
that
be
something
that
we'd
want
to
track
or
or
relate?
I
didn't
put
it
in
this,
but
it
does
make
sense
to
also
track
transaction
logs.
D
B
Right
now
would
no
would
like
would
it
make
sense
like
for
istio,
you
could
use
kiali
to
help
map
out
the
the
transaction
path.
We
were
looking
at
we're
working.
There's
a
webinar
yesterday
that
we
around
captain,
which
is
backed
by
dynatrace,
so
dynatrace
has
a
lot
of
that
transaction
logs.
There's
a
couple.
Other
tools
is
that
something
we
could
make
a
relationship
or
gather
that
information.
I
mean,
I'm
talking
really
high
level.
B
You
know
here's
the
url
to
go
if
you're
having
a
transaction
problem
or
if
you
want
to
look
at
the
transaction
for
this
service.
Here's
where
you
go
in
this
other
tool
to
go,
get
the
details.
It
doesn't
make
sense
for
us
to
try
to
replicate
all
those
other
tools.
It's
just
too
monumental
of
a
task.
Yeah.
D
It'd
be
being
able
to
talk
with
open
telemetry,
which
is
going
to
be
based
off
of
what
jaeger
jager
tracing
did
is,
which
dated
our
datadog
operates
off.
Of
that,
I
believe
dynatrace
has
open,
telemetry,
tizen,
tie-ins.
D
I
know
open
telemetry's,
gotten
more
popularity
as
jaeger's
given
up
gavin
over
some
of
the
pieces,
and
people
like
datadog
are
using
that
open
telemetry
standard
for
their
essentially
agent
or
collector.
D
D
C
Right
prometheus
is
predominantly
your
metric
storage
and
then
something
like
grafana
would
give
you
a
dashboard
in
front
of
that
and
grafana
can
deal
with
things
other
than
prometheus
as
a
back
end.
And
so
you
can,
you
know,
get
views
for
some
of
that
stuff
through
grafana,
but
it's
not
traditionally
what
communities.
G
B
Okay,
I'll
just
put
that
as
a
placeholder.
For
now
we'll
need
to
look
at.
If
there's
a
you
know.
Basically,
what
we
want
to
do
is
just
you
know,
provide
a
one-stop
shopping
for
this.
This
detail,
and
this
would
probably
be
along
those
same
lines
of
one
to
end,
because
I'll
put
that
in
there
it
may
some
of
these
may
sit
like
dynatrace
will
sit
across
all
clusters
and
stuff
like
that,
so,
and
look
from
environment
to
a
trace.
B
B
If
we're
going
to
have
this
relationship,
if
we
have
multiple
environments,
each
one
of
them
has
a
cluster,
then
every
cluster
needs
to
have
a
specific
key
value
pair
at
that
level
that
one
I
have
to
to
think
about
some
more.
C
B
And
then
I've
been
doing
some
research
on
container
image
dependencies
so
specifically
ignoring
you
know:
we're
deploying
a
ear
file
out
to
a
webster
server,
even
though
artilles
can
do
that,
I'm
kind
of
ignoring
that
and
focusing
on
the
microservice
world
around
containers.
B
So
one
of
the
things
I
found
on
this
front
is
being
able
to
hook
into
the
output
around
snick
there's
a
one
called
dependency
track
and
cyclone
dx
that
allow
us
to
go
ahead
and
grab
the
information.
Let
me
see
if
I
still
have
I
gotta
run
in
here.
B
There
it
is
so
some
of
the
the
data
that
I've
been
playing
with.
I
took
our
our
doxie,
which
is
based
on
a
golang
hugo
server
and
actually
tied
into
our
build
in
a
as
a
test
come
on
where's,
my
project.
B
Sorry
all
these
pop-ups
are
in
the
way
of
course,
zoom
gets
in
the
way.
Then
all
right,
so
one
of
the
things
I
found
was
the
a
way
to
get
the
the
cves
this
scan.
I
did
didn't
bring
in
the
license,
but
there's
another.
B
Basically,
during
your
build,
you
can
scan
the
basically
all
the
go
lane
modules
and
bring
those
into
the
licenses,
but
just
looking
at
this
level
of
what
we
have
and
being
able
to
bring
these
dependencies
in
which
version
the
license,
that's
associated
them
and
then
their
their
vulnerabilities
is
what
I'm
thinking
at
that
level
here.
B
And
I
don't
know
what
really
to
call
those
those
dependencies
right
now,
I'm
using
the
word
packages
you
know
in
the
node.js
I
believe
they're
referred
to
packages,
go
lane
their
modules
python.
What
are
they
modules?
You
know,
so
each
each
language
has
their
own.
Like
you
know,
at
the
os
level,
you
got
the
rpm
packages,
for
example,
so
how
to
describe
these
may
be
something
that
to
make
it
generic
enough
could
be
a
challenge,
but
the
idea
behind
it
is
be
able
to
go
through
and
grab
the
package.
B
So
some
of
the
other
information
I
was
looking
at
the
service
catalog
level
to
bring
in
as
part
of
that
now
the
kind
of
the
decision
point
on
this
is
we
can
actually
the
tools
that
scan
will
spit
out
a
json
file.
B
That's
either
spdx
format
or
basically,
it
looks
like
a
cyclone
dx
format
that
we
can
actually
load
this
data
and
render
save
the
data
in
our
in
in
ortilius
and
be
able
to
manage
it
ourselves
and
do
it
that
way
or
we
again
do
a
link
to
the
tools
as
part
of
that
process,
I'm
kind
of
leaning
to
where
we
can
do
a
link,
because
if
the,
if
somebody
runs
like
a
snake
or
something
like
that,
that
they
could
just
pass
us
the
url
and
we
don't
have
to
try
to
go,
consume
a
the
output
as
a
snick,
a
security
scan
and
bring
it
in
and
we're
separated
by
like
four
steps
in
the
build
in
the
process.
B
So
I
think
it'll
be
easier
just
to
pass
along
a
url
on
that
front.
As
part
of
that.
So
I
still
working
on
that.
One
and.
F
So
steve,
do
you
have
the
links
to
those
docs
in
the
in
this
doc.
B
To
the,
if
you
actually
go
to
the
our
general
discussion,
here's
all
the
links
to
the
tools.
F
And
your
document,
that
top
document
is
the
one
you're
referring
to
that
you're
working.
F
B
Just
some
last
minute
coding
changes
and
then
I'm
just
gonna
wrap
this
up
so
right
now
we
have
this
big
details
box
about
a
lot
of
the
service
catalog
data
and
what
I'm
proposing
is
to
start
splitting
it
out
into
smaller
boxes
as
part
of
that
on
the
ui
front.
Just
so,
we
can
kind
of
move
things
around
and
like
on
the
on
these,
where
we
have
like
a
one
to
n
we're
going
to
have
a
table
tables
in
the
microservice
on
the
ui
front.
B
F
Just
yeah,
I
think
that
will
be,
I
think,
putting
them
in
logical
groupings
of
information
will
be
helpful
because
otherwise,
right
now,
I'm
where
I'm
like.
Where
is
that
it's
getting
too
much.
B
So
some
of
the
some
of
the
new
tables
that
we'll
have
in
artelias
in
the
component
details
are
going
to
be
like
the
the
runtime
logs.
These
are
you
know
the
environment,
the
endpoint
and
there's
the
log
file
for
just
the
and
then
the
dependencies
along
those
lines,
as
well
so
you're
starting
to
create
issues.
B
I'm
going
to
start
to
make
sure
that
we
track
all
this
and
start
if
ever.
If
this
looks
good
for
everybody,
if
we
get
a
nod
to
go
ahead
with
it
I'll
start
going
in
and
putting
the
issues
in
and
on
what
needs
to
be
changes.
B
Yeah
and
they
answer
your
question
christopher
the
this
virus
table.
Basically,
it's
like
a
key
value
table,
so
the
nice
thing
about
it
is
we
can
keep
on
adding
whatever
we
want
to
it
without
having
to
change
the
so
there's
a
a
row
for
every
attribute.
Basically,
so
that's
one
thing
that
we
can
keep
on
adding
to
without
having
to
make
table
changes
on
that
front.
B
F
B
You
know
these
are
basically
the
com
columns
that
we
have
that
we
have
to
deal
with
so
they're
they're
not
complicated
table
changes
at
all.
B
So
let
me
go
back
to
the
architecture,
so
all
the
all
the
links
to
the
docs
and
I'll
be
creating
the
once.
We
all
say
yeah.
This
looks
good
to
go
I'll
start,
creating
the
the
issues
in
the
artillious
repo.
Most
of
the
coding
changes
will
be
java.
Now,
let
me
put
in
here
and
python.
B
Will
be
java,
javascript
python
and
some
database
changes
new
reports
and
tables
will
back
with
a
microservice,
probably
written
in
python.
That's
what
we've
been
doing
lately.
F
And
you
don't
what,
if
somebody
wants
to
write
and
go?
Is
that
is
that?
Okay,
if
it's
in
a
microservice,
yeah.
B
B
And
then
take
a
look
at
the
cyclone,
so
the
dependency
tracker,
the
ui
I
was
showing
you
with
the
the
cves
that
takes
information
or
gets
loaded
in
from
cyclone
dx.
Cyclone
dx
seems
to
be
like
one
of
the
best
and
the
most
accurate
from
what
I
could
tell
scanning
tools
that
I
found
that's
open,
source
and
it'll
kick
out
not
only
its
own
format
but
also
spdx
format.
That's
kind
of
more
of
a
spd-x
is
more
of
an
industry
standard
from
what
I
can
tell
it.
B
Spd-X
is
out
of
the
cncf
side
and
then
finally,
the
hugo
website,
so
the
web,
the
hugo
version
of
our
website
is
up
and
running
on
that
url.
B
So
we
need
to
do
to
do
some
work
on
that
front
on
styling
and
getting
the
content
moved
over
from
the
wordpress
one,
and
we
just
need
to
get
that
done
before
the
15th,
because
our
our
renewal
is
expiring
on
siteground
for
the
wordpress
version,
there's
the
repo
and
that's
all
I
have
for
today.
So
let's
go
ahead
and
I
know
like
tracy
said:
there's
a
few
new
people.
F
F
Good
and
then
we
have
two
individuals
who've
joined
this
call
that
have
some
pretty.
I
would
say
that
they're
heavy
hitters
in
the
kubernetes
world,
jim
stock
and
and
aaron
bruckner,
so
welcome.
Why
don't
you
jim?
Why
don't
you
introduce
yourself.
G
Hi,
I'm
jim
stock
and
I've
spent
a
little
while
doing
infrastructure
whatever
and
then
all
kinds
of
stuff,
with
infrastructure
measurement
monitoring,
diagnosis
whatever
and
then
I
started
getting
into
hadoop
yarn.
Apache
spark
kafka
environment
implemented
prometheus
there
and
then
did
it
in
a
kubernetes
private
cloud
at
general
motors.
So
I
am
reasonably.
E
G
Acquainted
with
kubernetes
and
that
stuff,
but
I've
just
done
measurement
problem
diagnosis
triage
all
of
that
stuff.
Operational
support
after
architecture
is
done
and
thrown
over
the
wall
for
a
long
time,
which
is
why
I
mentioned
a
couple
of
weeks
ago:
the
point
about
the
catalog
or
the
itsm
data
structure
to
understand
and
how
to
map
right
so
splunk.
Whatever
you
use
to
detect
a
log
file,
you
got
to
be
able
to
attract
that
to
something
and
as
you're
talking
about
key
value
pairs
and
all
this
stuff.
G
People
need
to
think
about
that.
Obviously,
you
guys
are
aware
of
that
right,
but
as
part
of
your
engineering
process
prior
to
deployment,
you
got
to
know.
How
are
you
going
to
create
those
relationships
and
and
connect.
G
Even
if
you
use
dynatrace,
then
trace
has
multiple
different
layers
that
go
fetch
disparate
data
and
how
you're
going
to
reconnect
that
and
then
make
it
useful
to
a
software
developer,
an
operations
person
running
a
command
center,
an
architect
to
understand
what
did
he
do
wrong
in
his
design?
You
know:
there's
lots
of
different
layers.
Everybody
needs
pretty
much
the
same
data,
maybe
graphed
differently,
but
the
relationships
of
what
does
this
data
mean
to
the
company
is,
I
think,
really
paramount.
B
Having
all
that
data
is
not
very
useful,
yeah
and
that's
one
thing
I
forgot
to
mention-
is
at
the
microservice
level
all
that
service
catalog
data,
we'll
roll
that
up
to
the
logical
version
of
the
application
as
well.
So
the
number
of
vulnerabilities,
the
licenses
that
are
being
consumed.
G
B
G
Yes,
yeah
or
application
sets
yep
yep,
and
so
what
lan
I
mean.
Logically,
we
all
can
see
here's
the
hierarchy
in
a
microservice
cloud,
kubernetes,
whatever
you
want
to
call
it
environment.
The
question
is
the
each
individual
corporation
has
to
agree
upon
a
certain
language
and
then
kind
of
follow
the
rules
with
that
right.
So
itsm
makes
that
really
simple,
because
once
it's
once
your
catalog
is
sort
of
named
and
you
understand
what
those
names
apply
to
then
you
just
start
mapping
stuff,
whether
it's
auto
discovery
or
it's
hard
coded
somewhere.
F
Thank
you
that
was
thank
you,
jim
everybody
will
learn
he's
for
your
stick
in
terms
of
trying
out
new
things
and
learning
new
things.
So
we're
glad
to
have
you
because
I
think,
there's
a
you
know.
We
have
kind
of
a
variety
of
people
here
and
having
more
people
on
the
operations
side
and
giving
us
feedback
will
be
super
helpful
and
aaron.
Aaron
is
the
same
way.
Aaron
is
another
heavy
hitter.
He
he
has.
His
resume
is
pretty
vast
in
terms
of
tools
and
stuff
he's
played
on
so
aaron.
D
Sure
so
yeah,
so
I'm
aaron
I've
been
dealing
with
microservices
and
helping
previously
was
that
safe
arm
on
the
claim
side,
which
is
the
busiest
side
I
like
to
tell
people,
because
that's
where
we
everything
gets
interacted
with
at
a
large
scale,
but
helping
to
find
and
build
what
that
microservice
ecosystem
looks
like
and
what
makes
sense
versus
the
traditional,
service-oriented
architecture.
D
So
I
spent
several
years
a
couple
years
doing
that,
and
then
I
really
kind
of
throughout
my
history,
of
working
on
service,
oriented
architectures
and
supporting
them
transitioned
into
really
getting
the
sre
practice
started
and
helping
people
understand
you
know
how
do
we
deploy?
How
do
we
architect?
D
Not
only
you
know,
hey,
we
got
to
be
multiple
places,
but
how
do
we
architect?
You
know?
What
are
our
clusters
look
like?
What
do
we
start
building
from
a
failover?
So
that's
why?
When,
when
you
say,
we
want
to
trace
something,
it's
like
well
could
be
one
of
n
right
and
based
on
what
type
of
reliability
we
need
for
the
cost
that
we're
willing
to
put
out
for
that
reliability
is
where
these
things
sit,
and
so
I
was
doing
sre
in
a
hybrid
cloud,
so
basically
across
at
the
time.
D
Basically,
the
time
was
aws
and
our
own
on-prem
cloud
and
really
starting
that
journey
and
showing
them
how
in
in-depth
kubernetes,
we
can
basically
do
a
bunch
of
things
so
jim
would
like
it.
I
was
actually
a
major
push
for
a
single
source
of
truth,
whether
it
was
your
logs,
whether
it
was
your
observable,
your
observability
data
right
how
your
services
are
running.
D
It
should
be
in
in
the
same
place,
regardless
whether
it's
spread
across
clouds.
So
that
way
we
can
actually
correlate
data
right
and
do
our
jobs,
it's
crazy,
but
you
know
sometimes
people
looked
at
me
like.
I
was
crazy.
It's
like
guys,
you
know
if
we
lose
tracking,
whether
it
crosses
a
boundary
is,
is
a
big
deal
that
and
so
an
american
express
so
to
afford
that
doing
many
things
in
the
sre
world.
D
Most
people
don't
necessarily
tie
us
or
even
secure
it
together,
but
they
are
very
closely
related
in
a
lot
of
ways,
because
sre
should
know
how
everything
runs
in
and
out,
and
then
security
asks
a
question
and
you're
like
no.
We
can't
do
it
that
way,
but
we
can
do
it
this
way.
D
So
I
spend
a
lot
of
time.
That's
what
I
spend
my
time
on
right
now
and
american
express
is
helping
secure
the
journey
to
their
hybrid
cloud,
their
journey
to
using
kubernetes
and
why
you
may
use
vanilla
over
some
of
the
other
ones,
because
they
may
put
limitations
on
there
right
and
it'll,
be
something
I'll,
probably
track
as
you
get
into
some
more
financial
tests
that
they're
going
to
want
to
be
able
to
track
not
just
some
micro
services.
D
But
how
are
we
controlling
the
communication
between
micro
services,
even
in
the
cluster
right
so
stuff,
like
istio
authorization
policies,
network
policies
so
on
and
so
forth
across
that
board,
because
they
have
to
answer
to
auditors
right
like
they
have
to
answer
to
the
government
in
and
of
itself,
which
is
a
whole
which
is
fun
and
challenging
at
the
same
time?
But
that's
the
big
background
I
bring
is
microservices
hybrid
cloud
and
all
the
fintech
fun
stuff
of
control
mechanisms
that,
whether
we
like
it
or
not,
has
is
what
we
have
to
do.
B
Yeah
yep
definitely
and
then
there's
one
other
name.
I
don't
recognize
zach.
Why
don't
you
go
ahead
and
introduce
yourself.
E
Hey
I'm
zach
jones
started
my
career
actually
in
a
manufacturing
facility,
designing
like
automated
inspection
systems,
so
a
lot
more
like
physical
controls,
oriented
type
stuff,
it's
always
been
interested
in
automation
and
then
that
eventually
led
me
into
an
sre
role
and
that's
pretty
much
what
I
do
now.
My
company
cloudfit.
E
We
predominantly
do
it
it's
more
onboarding
of
like
we'll
just
say,
government
organizations,
particularly
dod
into
the
cloud.
So
do
a
lot
of
work
in
you
know
mainly
azure,
so
azure
us
government,
but
then
you
know
their
secret
and
top
secret
clouds
beyond
that
and
azure
stacks.
So
a
you
know
totally
disconnected
instance,
so
yeah.
So
there's
that
onboarding
piece
of
just
to
the
cloud
in
general.
E
But
you
know,
as
I'm
sure
a
lot
of
you
already
know
the
the
government
tends
to
lag
behind
in
many
different
ways
when
it
comes
to
technology,
so
kind
of
the
the
second
hat
I
wear
is
you
know,
kind
of
introducing
these
entities
to
really
just
kubernetes
in
general
or
containers.
You
know
a
lot
of
them.
E
Just
you
know
haven't
even
had
experience
with
docker
so
kind
of
showing
them
the
power
of
the
possible
so
to
speak
and
and
the
real
goal
there
is
to
you
know,
expose
them
to
you
know,
methodologies
and
practices
that
you
know
you
can
build
a
pipeline
with
so
to
speak
and
that
pipeline
can
have
an
authority
to
operate
in
these
restricted
environments.
E
And
then
all
code
push
through
can
kind
of
be
grandfathered
into
an
authority
to
operate.
So
we
call
that
continuous
ato,
but
it's
it's
basically.
The
whole
purpose
of
it
is
to
you
know,
make
it
to
where
necessary,
changes
to
like,
let's
say
jets
or
weapon
systems
or
whatever
don't
take
months
to
get
out.
They
can
get
out.
You
know
as
soon
as
we
need
them,
so
that's
kind
of
where
I'm
at
just
a
lot
of
work
in
secure
environments
and
yeah
happy
to
be
here
and
yeah.
B
Everybody-
and
I
did
put
the
link
to
our
discord
channel
if
you
guys
want
to
join
that,
there's
a
couple
categories:
there's
the
general
one,
then
the
dev
channel,
for
where
we
take
it
a
little
more
deeper
on
discussions
and
we're
on
the
discord
channels.
It's
just
not
limited
to
ortelius.
B
F
I
only
it
makes
it
interesting,
because
a
lot
of
you
know
spring
they
have
their
doesn't
have
a
registry
at
the
cluster
level
for
apis.
E
Good
question,
so
I
actually
have
no
idea
when
I
was
working
with
spring.
It
was
right
after
that.
How
was
it
it
was
that
huge
equifax
security
thing
that
was
related.
F
E
F
D
I
don't
know
spring
does
or
spring
boot
does
by
itself,
because
it
just
creates
a
really
a
memory
standpoint
through
annotations
within
side
of
it.
It'd
have
to
be
a
different
deployment
right,
so
spring
works
with
a
version
of
control
memory
space
on
top
of
the
jvm.
D
B
That's
a
netflix
and
I
can't
remember
if
it's
a
pub
sub
type
of
or
if
it's
auto
discovery
of
the
services.
I
think
it's.
I
think
eureka
does
auto
discovery
of
the
services,
and
I
don't
remember
if
there's
spring
services
or
any
type
of
service
in
the
java
world.
H
Yeah,
they
just
call
hook
from
git,
and
then
they
know
where
to
publish
to
services,
queries
they
register,
and
then
you
have
a
centralized,
a
gateway
that
manages
all
the
redirections.
So
far.
B
D
Right
right,
yeah,
but
most
your
major
wants
to
have
crds
and
kubernetes
so
like
glue
an
ambassador
have
crds
for
kubernetes
that
you
would
define
your
entry
into
the
proxy
and
it's
all
managed
through
that
way.
H
Yeah,
usually,
they
overlap
when
you're
bringing
spring
boot
applications
to
kubernetes.
Always
there
is
a
thing
when
you're
start
talking
about
services-
and
always
the
netflix
oss
is,
is
a
thing
because
they
do
more
things,
but
it
still
overlaps
with
kubernetes.
So
usually
we
we
just
move
out.
We
we
just
with
using
this
discovery,
surgeon
stuff
that
springboat
uses,
because
it's
not
really
powerful
on
curenet.
F
Yeah,
well,
I
get
a
lot
of
questions
on
that
in
particular,
and
if
there's
any
integration
we're
doing
at
the
pulling
their
transactions
from
the
api
gateway,
I've
not
gone
down
that
road
to
think
about
how
to
do
that,
but
I
bring
it
up,
so
we
can
think
about
if
that's
something
we
wanted,
but
it
is
language
specific
and,
on
another
note
sergio.
F
So
I
think
that,
if
you
want
to,
if
you
have
something
you
can
send
me
and
let
me
play
with
it
and
see
what
it
looks
like
post
it
around,
I'm
happy
to
do
that.
But
I
I
I
think
the
gold
background
on
the
evil
lord
would
be
as
good,
and
we
might
only
need
that
badge
and
have
different
names
on
it.
H
Yeah,
I
was
thinking
like
I
think
it's
good
good
enough,
like
maybe
just
changing
background,
is
going
to
be
enough
on
the
name
because
yeah,
it's
look.
Okay,
other
way,
we
are
going
to
like
thinking
about
like
how
to
modify
the
the
appearance
of
this
guy.
Every
time
we
need
to
to
relate
it
to
some
kind
of
level
or
collaboration.
F
F
F
Okay,
all
right
everybody,
huge
thanks
and
on
one
final
note
we
did
have
a
really
interesting,
cdf
meetup
yesterday
with
captain
I
will
post
it
when
I
get
it
uploaded,
I
will
post
it
out
on
the
discord
channel.
F
It's
really
interesting
in
how
kept
in
is
changing
the
ci
cd
world
into
something
that
has
a
kind
of
a
control
plane
listener
to
fire
off
events,
and
it's
something
I
think
we
as
we
watch
ci
cd
morph
into
something
more
intelligent,
that
we're
probably
going
to
see
more
tools
like
that,
and
it
could
be
something
we
could.