►
From YouTube: Keynote: An update on the extremely boring and uninteresting world of Linkerd- William Morgan
Description
Keynote: An update on the extremely boring and uninteresting world of Linkerd- William Morgan, Buoyant
In this keynote, William Morgan, CEO of Buoyant and one of the creators of Linkerd, will deliver a project update on the extremely boring world of the Linkerd service mesh, the CNCF’s only graduated service mesh. William will cover all the uninteresting things happening in this boring project and discuss some of its profoundly non-exciting approaches to some perfectly ordinary challenges.
A
All
right:
well,
we
made
it
thanks,
I'm
sure,
there's
a
service
mesh
joke
in
there
somewhere.
First
of
all
giant.
Thank
you
to
craig
for
that
extremely
musical
introduction.
I
never
thought
I
would
see
that,
but
now
I'm
glad.
I
think
that
I
have
so.
My
name
is
william
morgan.
I'm
one
of
the
creators
of
linker
d,
I'm
ceo
of
a
company
called
buoyant.
A
A
Our
adoption
grew
by
over
double
in
in
europe
and
overall
across
the
world
since
2021.,
if
you
were
virtually
attending
kubecon
last
year,
the
linker
d
team
keynoted
on
all
the
ways
that
linkerity
has
been
used
to
combat
the
global
coving
19
pandemic.
I'm
very
proud
of
that.
We
added
fuzz
testing
to
the
project.
A
Okay,
all
right,
so
it's
also
going
to
be
a
very
non-boring
week.
If
you
are
a
link
or
d
enthusiast
here
at
kubecon,
we've
got
a
whole
bunch
of
talks.
You
actually
should
have
a
piece
of
paper
that
will
list
a
bunch
of
them
too.
Two
I'll
point
out
here
later
in
this
conference,
linker
d,
maintainer,
kevin
linecooler,
is
going
to
talk
about
how
linkrd
achieves
some
of
its
zero
config
powers
and
then
on
thursday
service
mesh
at
scale.
A
You
can
learn
about
how
xbox
cloud
gaming
uses
linker
d
to
secure
22
000
pods
around
the
world,
but
of
course
all
these
talks
are
really
exciting
and
I
recommend
that
you
join
them.
So,
speaking
of
exciting,
we
are
here
because
we
think
the
service
mesh
is
exciting.
At
least
I
hope
that's
why
you're
here
you
know.
A
Otherwise
what
are
you
doing
with
your
life
and
it's
natural
for
us
to
talk
about
how
exciting
the
service
mesh
is
and
what
you're
going
to
hear
the
rest
of
the
day
is
about
these
exciting
features
and
the
exciting
new
use
cases
and
all
of
the
fun
stuff
we
get
to
do,
and
that's?
Okay,
you
know:
that's
fine,
like
we
can
geek
out
about
service
meshes,
because
we
are
implementers
and
we
are
enthusiasts.
We
can
geek
out
about
performance
right.
A
We
can
geek
out
about
resource
consumption
and,
most
importantly,
we
can
geek
out
about
all
of
the
cool,
crazy
things
that
we
get
to
add
to
the
service
mesh,
and
I
don't
know
where
I
found
this
image,
but
it
really
is
remarkable
if
you,
if
you
stare
at
it
for
a
while
right,
but
our
service
mesh
adopters,
don't
actually
want
the
service
mesh
to
be
exciting.
They
want
it
to
be
boring
right.
A
We
might
like
the
excitement,
but
if
you
are
the
poor
soul
who
has
to
actually
operate
the
service
mesh,
you
want
to
be
boring.
Boring
means
you
don't
want
it
to
have
surprises
you
don't
want
it
to.
You
know,
suddenly
do
something
unexpected,
that's
never
good
boring
means
you
don't
want
to
have
to
go
through
a
bunch
of
scary
tasks,
to
keep
it
up
to
date
or
to
maintain
it
or
to
keep
it
healthy.
Right
and
boring
means
that
you
don't
want
to
have
to
wake
up
at
three
in
the
morning.
A
Thinking
about
is
my
service
mesh.
Okay
right
do
I
have
to
hire
someone
to
deal
with
this
thing.
That's
the
opposite!
Okay,
and
we
have
a
very
particular
strategy
in
the
linker
d
project
for
how
we
address
that
and
how
we
make
things
boring,
and
that
is
you
know
we
make
it
simple
right:
that's
that's
our
that's
our
secret
right
and-
and
we
talk
specifically
about
operational
simplicity.
So
as
you
you,
the
adopter
of
the
service
makes
you
the
operator.
What
does
it
mean
for
linker
d
to
be
simple?
A
Well,
one
thing
is,
it
means
lingerie's
observable.
Every
aspect
of
linkage
is
something
that
you
should
be
able
to
see
and
measure
and
record
right.
It
means
linkedin's
got
to
be
understandable.
You
should
be
able
to
break
it
down
into
its
little
components
and
you
should
be
able
to
understand
how
each
one
works
and
the
documentation
should
be
clear
and
like
that
design
should
make
sense.
And
finally,
lingerie
should
be
predictable
right.
A
If
you
expect
it
to
do
something,
it
should
do
that
thing
and
if
you
expect
it
to
never
do
something,
it
should
never
do
that
thing
right
and
so
here's
our
here's,
our
top
secret.
I
can't
believe
I'm
giving
this
information
away.
Here's
our
top
secret
design
philosophy
number
one,
do
no
harm
right.
So
if
you
have
an
application
that
is
running
and
you
add
linkedin
to
it,
the
application
should
continue
running
sounds
incredible,
but
that
is
actually
quite
hard
to
achieve.
A
Second,
you
want
to
be
kubernetes
right,
so
we're
all
adopting
kubernetes,
whether
we
like
it
or
not,
we
are
stuck
using
kubernetes.
Linker
d
should
feel
like
kubernetes.
It
should
smell
like
it.
It
should
be
as
close
to
kubernetes
as
possible.
Third
make
the
data
plane.
Awesome
right
like
this
is
something
very
unique
to
linker
d.
I'd
love
to
talk
a
lot
more
about
this,
but
this
is
a
big
part
of
of
linkery's
goal
of
operational
simplicity.
Of
course.
Fourth,
we
want
to
be
modular
and
composable.
A
If
you
go
to
the
workshop,
that
is,
I
think,
wednesday,
about
how
we
do
cross-cluster
failover.
I
think
it's
a
really
nice
example
of
how
we
came
up
with
a
really
modular
architecture
there.
Finally,
buoyant
has
introduced
fully
managed
linkery.
This
is
our
ability
to
basically
treat
linguity
as
a
managed
service.
I'm
really
excited
about
this.
I
would
also
love
to
tell
you
a
lot
more
about
that.
We're
hiring.
You
can
find
me
at
the
buoyant
booth
out
there
or
at
the
buoyant
booth
in
kubecon.