►
From YouTube: GitLab CEO - Microservices and Architecture at GitLab
A
B
Thanks
for
doing
this
and
I
think
the
trend
is
that
we're
getting
more
realistic
about
micro
services
a
couple
years
ago,
everyone
was
like
everything
should
be
as
small
as
possible.
Let's
take
our
big,
app
and
split
up
in
500
different
micro
services,
I
think
industry,
seeing
that
that
that
doesn't
work,
but
there's
there's
a
clear
benefit
to
micro
services
like
operational
reliability
and
things
like
that.
So
it
is
now
getting
more
popular
with,
but
without
the
extreme
kind
of
patterns
we
saw
earlier
so
we're
getting
pragmatic
about
microservices
and.
A
A
B
B
You
could
argue
that
that
is
something
that
should
be
highly
performant
very
fast,
and
maybe
that
should
be
run
somewhere
else
now,
I
think
right
now
the
bottleneck
for
that
process
is
probably
the
database
more
than
the
service
and
I.
Sorry
I,
don't
have
any
in-depth
knowledge,
but
I
could
I
could
see
cases
for
things
like
that
that
are
very
attractive.
B
A
Make
sense
so
a
lot
of
people
that
get
led
probably
have
the
understanding
they're
introducing
more
services
at
gate
lab
is
something
that's
been
very
much
discouraged
from
leadership
for
a
long
time,
and
people
might
have
the
impression
that
you
would
kind
of
in
general,
just
be
blanket
opposed
to
this
change,
but
I
wonder.
Instead
what
you
think
about
how
you
might
like
to
see
people
make
new
suggestions
about
introducing
services
into
the
architecture
of
breaking
parts
of
the
monolith.
Our
yeah.
B
B
The
product
managers
can
opine
and
probably
they'll
call
in
a
few
of
our
fellow
and
staff
engineers
to
give
an
opinion
but
I
one
thing:
we
always
do
at
gate
lab
as
we
iterate
so
splitting
doing
something,
that's
a
whole
ton
of
work
and
it
requires
everyone
else
to
hold
while
you're
doing.
That
is
probably
not
a
good
idea,
so
I
probably
make
it
smallest,
let's
not
say
micro,
but
the
smallest
possible
proposal.
Then
it
can
be
and
then
build
on
that
success
to
go
for
it.