►
Description
An Update for May has been Posted here: https://youtu.be/n0aMc9MFQtU
This video is to answer a few questions about the current state of tracking CI/CD minutes in public projects.
A few of the issues referenced are
- Fixing a bug displaying the wrong month for usage: https://gitlab.com/gitlab-org/gitlab/-/issues/357909
- Adding Shared Runner Duration to user namespaces: https://gitlab.com/gitlab-org/gitlab/-/issues/345795/
This work is being tracked in this epic where you can leave comments or in the existing issues: https://gitlab.com/groups/gitlab-org/-/epics/7561
A
And
james
I'm
a
product
manager
for
the
pipeline
execution
group
in
the
verified
stage
gitlab
today.
I
want
to
walk
through
where
we're
at
progress
wise
in
making
shared
runner
minutes
visible
and
how
you
can
equate
that
to
your
shared
ci
or
your
cicd
minutes,
as
well
as
we're
making
that
transition
to
count
and
enforce
the
limits
with
public
projects
for
any
of
the
tiers
free
premium
or
gold
on
gitlab.com.
A
So
we'll
start
here,
so
this
is
my
personal
account,
so
it's
all
public
projects,
it's
on
free
tier,
just
to
make
sure
that
we're
getting
a
good
experience
here
and
what
we've
enabled
today.
So
this
is
within
my
user
preferences
and
then
the
usage
quota.
So
I
can
see
my
ci
minutes,
so
I
can
see
my
cn
minutes
by
month.
A
This
is
for
all
of
the
projects
that
aren't
within
a
group
and
that's
where
there's
a
little
nuance
of
how
we're
counting
these
minutes
and
what
we're
displaying
into
the
different
places.
So
you
can
see
here.
I
have
a
couple
of
projects
that
I've
been
running,
some
pipelines
on
and
getting
some
minutes
accumulated
and
you
can
see
that
breakdown
here.
There
is
a
known
bug
where
we're
showing
the
wrong
month
on
this
chart
today.
That
bug
is
scheduled
to
be
fixed
in
the
15.1
milestone
which
releases
in
june.
A
That
being
said,
this
data
is
accurate
if
off
by
a
month,
not
a
great
user
experience
and
we're
really
sorry
about
that,
but
the
data
that
is
showing
in
march
right
now
is
accurate
for
april
we're
just
off
by
a
month.
A
I
did
want
to
call
out
that
the
big
difference-
the
thing
that's
happening
on
june
1st-
is
that
this
enforcement
is
going
to
happen
on
all
public
projects,
regardless
of
creation
right
now,
we're
counting
those
minutes
for
any
project
that
was
created
after
july
17th
of
2021,
and
so
here
you
can
see
my
jh
java
example
project.
A
My
test
minutes
my
test
nets,
private
those
are
all
projects
that
I've
created
since
then,
this
blog
site
jhambach.gitlab.io
that
one
I
just
ran
a
pipeline
on
about
15
minutes
ago
that
ran
for
two
and
a
half
minutes.
But
you
see
here
that
there's
no
minutes
being
counted,
that's
because
that
project
was
created
prior
to
that
date.
So
after
we
make
the
flip
or
switch
flip
that
switch
on
june
1st,
then
those
minutes
will
start
to
accumulate
and
they'll
count
against
my
quota
of
400
minutes.
A
A
A
Same
bug
persists
here,
we're
off
by
a
month,
but
this
data
is
accurate,
so
I'm
showing
that
I've
used
17
ci
cd
minutes.
You
can
see
here
that
that
really
adds
only
up
to
15.
there's
a
little
bit
of
rounding,
that's
going
to
show
up
as
a
problem,
so
I'll
file
a
bug
against
that
to
make
sure
that
that
is
accurate
and
what
you're
seeing
is
what
we're
counting
better,
at
least
within
the
tolerance.
The
other
difference
here
is
the
shared
runner
usage.
A
So
what
we're
displaying
right
now
on
next
is
how
many
actual
runner
minutes
were
used.
So
this
is
how
long
the
runners
were
running
for
those
pipelines,
and
this
shows
28.67
so
there's
a
little
bit
of
a
difference
there,
and
so
the
reason
for
that
is
cost
factor.
So
I'm
showing
17
ci
cd
minutes
here
and
20,
let's
say:
26
min
28
minutes
for
shared
runner
duration,
and
so
I'm
going
to
just
pause
and
find
the
documentation
page
for
that.
Real
quick
and
to
better
explain
what
this
quota
means.
A
A
So,
as
you
can
see,
the
job
duration
for
those
public
projects,
you'll
multiply
that
times,
0.08
and
that's
how
many
cicd
minutes
will
actually
be
used
and
just
to
make
sure
sorry
point:
zero.
Zero,
eight,
not
point
zero,
eight,
so
for
125
minutes
of
job
time,
you're
using
one
ci
cd
minute
in
a
public
project
and
that'll
be
for
all
public
projects
after
june
1st
for
all
internal
and
private
projects,
regardless
of
when
they
were
created,
and
regardless
of
your
tier,
it's
a
one-to-one.
A
So
there's
one
cost
factor
there
so
for
every
one
minute
of
job
duration,
you're
going
to
have
one
minute
of
ci
cd
minute
quota
used,
and
so
that's
why
that's
a
little
bit
a
little
bit
of
an
odd
experience
and
what
you're
seeing
between
shared
rendered
usage
and
the
pipelines
today?
Why
there's
not
an
exact
match?
So
the
shared
runner
usage-
you
see,
doesn't
have
these
extra
graphs
underneath,
so
you
can
actually
tell
which
projects
are
using
up
those
or
using
that
shared
runner
duration.
A
So
the
first
is
the
bug
that
I
called
out
the
wrong
month
on
there.
That
is
scheduled
to
be
fixed
in
15-1,
which
will
roll
out
in
june.
As
soon
as
that's
available.
On.Com,
though
you
don't
have
to
wait
until
june
22nd
to
see
that
so
that
will
just
flip
as
soon
as
that
bug
is
resolved,
and
then
the
other
is
providing
that
visibility
for
the
shared
render
usage
for
user
name
spaces.
A
A
So
you
can
start
to
better
track
what
projects
are
using
up
that
duration
on
the
shared
runners
and
make
adjustments
as
you
go,
there's
a
couple
other
usability
improvements
that
I
want
to
show
here,
we're
breaking
out
the
drop
down
so
that
you
can
have
month
and
year.
So
it's
not
a
really
long
one.
A
If
you
have
a
lot
of
history
of
your
projects
and
for
users
who
have
a
lot
of
projects
right
now,
this
table
is
honestly
not
really
very
useful,
so
we're
going
to
start
to
sort
that
by
your
shared
render
usage
or
whether
your
ci
cd
minutes
used.
So
your
top
projects
are
at
the
top.
You
can
identify
what's
happening.
You
know,
what's
really
using
all
of
those
minutes
without
having
to
scroll
through
you
know,
potentially
dozens
or
even
hundreds
of
pages,
depending
on
the
number
of
projects
that
you
have,
if
they're
all
running
pipelines.
A
So
those
are
a
couple
of
improvements
and
where
we
stand
today,
so
I'm
going
to
post
this
video
as
a
response
to
a
couple
of
comments,
but
I'll
share
it
on
our
gitlab,
unfiltered
and
I'll
link
to
all
of
these
issues.
As
well,
if
you
want
to
continue
the
discussion
about
where
we're
going
with
this
project,
there
thanks.