►
From YouTube: Monitor:APM Weekly Meeting 2019-11-27
Description
Weekly meeting for the Monitor:APM team. Discussions include:
- How to handle the various issues for logging
- Triage
A
B
B
B
Which
again
you
created,
there
is
a
log
small
issue.
There
is
a
filtering
issue
and
there
is
an
epic
now,
that's
it's
good
I
mean
I've,
read
all
of
them,
but
I'm
trying
to
consolidate
everything
into
a
single
one.
So
what
I'm
proposing
is
that
I
think
we
are
done
with
the
discussion?
I
think
that
we
agreed
that
we
are
going
to
walk
on
the
existing
terminal
view
for
the
log
Explorer
and
add
those
capabilities
to
that
view.
B
Okay,
there
is
an
ongoing
flow
about
whether
we
should
use
the
new
design
or
continue
working
on
the
town
of
view
and
I.
Think
Edwin.
You
pointed
out
that
it
will
save
us
a
lot
of
time
if
we
continue
to
work
on
the
terminal
view,
while
still
using
logs
from
elasticsearch.
So
I
fully
understand
that
and
I
think
this
is
the
right
approach
for
now.
B
B
Walking
and
enhancing
this
view,
for
instance,
adding
filters
creating
a
search
bar,
refreshing,
the
page
and
more
so
in
order
for
me
to
understand
where
we
stand
and
I
really
don't
want
to
interfere
with
with
the
work
I
just
want
to
make
sure
we
are
working
on
the
right
priorities.
I
closed
all
the
issues.
Okay,
discussions
are
important
and
I've
opened.
B
A
meta
issue
where
we
need
to
do
is
just
to
check
box
once
we
are
done
with
one
of
the
one
of
the
tasks
and
if
we
need
to
add
more
tasks,
we
can
add
more
tasks.
I
don't
want
to
I
want
to
stop
any
any
work
that
is
being
done
because
of
any
process.
So
I
want
to
understand
that
when,
if
that's
okay
with
you,
because
I
think
you're
doing
the
heavy
lifting
there
and
so
I
just
want
to
make
sure
we
can
continue
with
that
approach,
sure.
C
B
Okay,
so
once
we
are
done
with
that
issue,
every
like
every
amount
we
can
just
you
can
just
check
box
and
say:
okay
mark
done,
and
if
we
need
to
open
a
separate
issue,
I'll
be
more
than
happy
to
do.
That
will
be
familiar,
did
one
for
the
time
picker.
So
if
you
think
that
there
should
be
a
discussion,
maybe
you
should
read
a
comment:
let's
open
an
issue,
an
open
issue-
and
let's
have
this
discussion
on
that
issue,
so
we'll
keep
this
meta
issue
just
to
tracking
keep
track
the
progress
of
this
project.
Okay,.
B
We
need
to
open
issue.
Let's
just
put
a
comment:
that's
open
any
issue
and
I'll
open
an
issue.
I
put
a
link,
I'll
put
a
link
in
the
description.
Let's
have
this
conversation
in
the
issue
because
those
issues
description
tend
to
you
know.
Multiple
conversation
happening
and
I
want
to
make
sure
the
dislocation
only
track
the
progress
of
what
we
are
doing.
B
B
We
are
starting
to
start
to
experiment
with
something
else
which
we
call
the
user
journey
of
the
user
of
the
workflow.
And
if
you
look
at
our
monitoring
vision
page,
you
will
see
that
we
have
listed
five
different
user
journeys.
We
call
them
user
flow,
it's
instrument,
triage,
improve,
resolve
and
alerting.
B
Okay,
the
reason
I'm
telling
you
all
of
that
is
that
there
is
a
special
interest
of
the
triage
workflow,
okay
and
in
this
epoch,
I'll
kind
of
try
to
explain
what
it
what
triage
means,
what
tears
mean
for
us,
so
for
us
triage
mean
and
as
I
mentioned
in
this
background,
this
is
the
entire
process
of
detecting
and
identifying
a
problem
in
your
application.
Okay
in
some,
if
you
look
to
the
how
the
industry
we
find
triage,
should
really
just
assign
the
right
person
to
the
problem.
B
So
here
in
triage,
we
kind
of
clubbed
together,
the
the
triage
in
the
hood
cause
analysis.
Okay,
in
some
cases
there
is
a
triage,
workflow
and
root
cause
analysis.
The
reason
I'm
explaining
all
of
that
is
that
we
would
like
to
take
this
end-to-end
flow
of
triage
and
see
what
we
need
to
do
in
order
to
bring
it
to
two
minimum
to
mature
it
to
a
minimal
level.
B
So
we
as
well
still
walking
on
category
basis-
and
we
will
still
work
on
that,
but
in
the
coming
future,
probably
in
the
next
iteration,
we
will
start
to
think
about.
How
can
we
take
this
flow
now
go
over
this
flow
and
understand
what
we
need
to
do
on
our
product
in
order
to
make
sure
that
we'll
have
a
streamline
experience
for
our
users
to
complete
the
slash
flow,
and
this
is
like
the
minimal
the
minimal
things
that
the
user
needs
to
do
now.
B
B
So
if
we
look
at
data,
if
we
look
at
any
other
companies
that
are
in
this
space,
we
can
see
that
most
of
the
time
the
triage
flow
starts
with
an
alert
and
in
that
a
lot
from
that
alert,
you
immediately
can
go
to
the
right
view
and
start
to
do
the
troubleshooting
on
the
root
cause
analysis
going
from
one
view
to
another.
So
we
simplify
the
workflow
that
we
start
with
an
alert.
B
The
metric
and
the
threshold
that
was
caused
and
formed
that
child
should
be
able
to
drill
into
the
long
Explorer
and
from
there
we
should
have
some
sort
of
a
link
to
old
traces,
so
the
user
would
be
able
to
see
which,
which
request
caused
that
problem
and
what
is
the
user?
What
is
the
user
impact
for
the
user
experience
impact?
B
So
this
is
the
flow
it's
it's
pretty.
We
kept
it
pretty
minimal
and
if
we
look
at
the
epic
and
the
epic
explained
the
user
journey,
so
first
of
all,
as
I
mentioned,
we
need
to
get
a
notification.
Then
we
need
to
look
at
the
view
of
the
metric
chart
in
the
incident
itself
and
then
to
drill
down
to
the
log
Explorer
to
preserve
the
same
context.
B
B
But
if
you
will
go
over
this
this
workflow
and
try
to
identify
if
there
are
any
gaps
that
you
think
think
are
missing,
I
think
that
we
kind
of
keep
it
keep
it
minimal
and
most
of
the
things
are
things
that
we
are
walking
on
or
frog
start
working
on
in
the
next
iteration.
But
you
know,
take
your
time
go
over
it
and
you
know
in
the
if.
A
A
About
making
make
sure
that
is,
we
have
a
an
incident,
a
metric
that
creates
an
incident
that
you
know.
We
look
through
the
log
explore
and
traces
to
make
sure
we
can
carry
the
context
of
which
machine
which
time
frame
which,
whatever
through
that
I,
think
that
we've
got
a
unique
ability
to
integrate.
Since
we've
got
all
that
information
yeah.
B
Yeah
and
I
mean
this
is
one
this
one
experiment
that
we
are
doing
and
if
we
submit
is
successful
and
we'll
get
more
feedback
from
our
users.
We
will
both
enhance
this
flow
and
multiple
multiple
workflows
to
tackle
the
triage
flow,
because
you
know
we
start
with
only
one
alert
to
go
to
one
view
and
then
another
view
may
be.
There
are
other
ways
for
users
to
do
triaging.