►
From YouTube: 2021 02 16 Database Team Meeting
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).
A
A
One
of
them
is
in
our
bucket
this
query:
performance
review
requested
and
andreas
assigned
it
to
himself.
Today
the
rest
of
them
are
assigned
to
teams.
Three
of
them
are
going
to
continuous
integration.
One
of
them
is
going
to
group
runner.
Is
there
anything
else
outstanding
that
we
should
prioritize
jose?
Is
there
anything
you're
waiting
on
us
for.
B
B
Additionally,
I'm
creating
some
I'm
working
at
the
moment
in
our
own
book,
trying
to
explain
how
to
gather
all
the
data
that
we
need
for
the
issues,
and
I
would
like
to
kindly
ask
for
your
revision
as
well
your
review
for
the
team
to
add
some
values
there.
As
soon
as
I
have
the
initial.
Mr,
it
sounds
good
to
everyone.
A
Yeah
is
this
regarding
the
the
replication
lag
that
issue.
C
A
All
right,
if
there's
nothing
else,
you
can
probably
drop
off.
Okay,.
A
E
Yeah,
just
I
know
there
we
enabled
re-indexing
today
in
the
morning
and
it
worked
out
fine,
so
no
issues
and
there's
a
couple
of
links
there.
I
didn't
know
about
the
sre
newsletter,
which
was
really
useful
to
see
yesterday.
E
So
this
is
a
way
of
sort
of
letting
mostly
the
encore,
know,
I
think,
or
maybe
also
sre
more
worldly,
but
basically
whenever
there
is
an
issue
that
needs
to
be
passed
along
to
the
to
the
upcoming
on
calls,
it
makes
sense
to
post
on
those
sre
newsletter
issues,
something
I
didn't
know
before
and
then
yeah.
E
This
is
happening
on
on
weekdays,
so
the
re-indexing
happens
on
weekdays
kicks
off
at
6am
in
the
morning
today
it
took
about
four
hours
because
it
was
hitting
very
large
indexes
and
it
also
freed
up
80
gigabytes
of
space
today.
So
I
would
expect
that
the
timing
goes
down
a
bit
once
we
hit
smaller
tables.
So
today
was
all
about.
Ci
builds
indexes
and
they
are
pretty
large.
E
All
the
tables
are
pretty
large
and
once
we're
done
with
them,
maybe
the
the
timing
should
should
go
down
a
bit
at
least
yeah.
That's
that's.
You
know
just
as
a
heads
up
so
this
is
this
is
happening.
A
Automatically
from
now
on
is
there
is
this
epic
gonna
remain
open.
E
Right,
let
me
take
a
look
through
those
issues.
Maybe
maybe
we
can
just
close
them,
but
I'll
take
a
look
and
close
the
ones
that
we
don't
need.
D
As
a
sort
of
general
observation,
I
spoke
about
that
like
half
an
hour
ago
with
the
library
team.
Personally,
I
think
it's
always
good
to
actually
close
out
a
large
epic,
even
if
there
are
some
like
sort
of
small
things
left
in
the
backlog
and
open
up
a
follow-up,
epic
and
say
like
further
improvements.
Something
else
right
so
that
you
actually
feel
like
this
thing
is
delivered.
We've
closed.
D
It,
maybe
there's
some
other
stuff
for
the
future,
but
I
think
that
makes
it
a
lot
easier
than
like
having
these
sort
of
95
epics,
where
actually
everything
is
kind
of
done,
and
then
they'll
stay
in
that
state
for
like
who
knows
how
long
right.
So
I
like
that.
D
A
E
Yes,
so
this
is
about
the
testing
pipeline
and
we're
basically
now
reporting
the
the
total
database
size
change
as
a
metric
migration,
which
is
kind
of
nice.
You
can
see,
you
know
you
can
immediately
see
how
that
how
large
it
was
in
production
or
your
dropper
table
and
then
similar
things
so
that's
happening,
and
then
the
other
one
prerequisite
basically
a
bit
of
a
funding
for
adding
what
we
call
migration
observers.
E
So
we
can,
you
know
we
continue
to
add
different
types
of
observers
for
immigrations,
which
report
different
statistics.
One
of
them
is
total
database
size,
I'm
also
working
on
query
statistics,
and
that
was
a
bit
of
a
foundational
issue
and
yep.
Those
are
fun.
C
Three
yeah,
the
first
one
that
was
just
a
customer,
reported
issue
with
the
audit
events.
Migration,
that's
I
think
we
have
worked
through
that
finally
and
fixed
it,
and
the
customer
is
able
to
upgrade
so
that
one
is
done.
C
D
C
One,
the
second
one
there's
is
the
same
thing.
It
was
a,
I
think,
the
audit
events
primary
key
had
the
wrong
name
and
we
were
fixed
that
and
yeah.
The
last
one
was
fixing
that
helper
that
wasn't
properly
syncing
the
data
when
the
column
was
renamed,
but
that.
E
Yes,
so
statistics
for
database
migrations-
this
is
this
is
about
adding
information
from
pg
start
statements.
So
basically
you
get
a
nice
table
of
normalized
queries
and
their
statistics
and
for
each
migration
this
is
practically
done.
I
think
so.
I'm
handing
this
to
review
today
and
I'm
basically
using
those
other
two
issues.
So
the.
E
They
both
have
migrations,
so
I'm
basically
using
them
to
test
the
testing
pipeline
to
see
some
results,
so
I'm
not
going
to
close
them
or
I'm
not
actively
working
on
them,
but
they're,
basically
an
example
for
migrations
that
are
easy
to
test
or
that
give
interesting
results.
C
Yeah,
I
don't
really
have
much
to
report.
I
wasn't
able
to
really
accomplish
a
whole
lot
last
week.
I
don't
know
how
much
I'll
be
able
to
accomplish
this
week.
Really
I'm
I
go
tomorrow
for
my
mri
and
then
I
have
my
follow-up
on
monday
with
a
doctor,
so
hopefully
by
then
I
will
have
a
better
idea.
What's
going
on.
A
All
right
is
there
anything,
that's
not
on
this
board.
That's
being
worked
on
right
now,.
A
Well,
the
brief
silence
means
nothing
top
of
mind
so
jump
over
to
the
triage
board
and
if
a
memory
serves
this
goes
from
giannis
to
andreas
this
week.
Yep,
yes,.
C
Yet
if
there's
anything
you
want
me
to
look
at
since
I'm
not
really
the
most
the
greatest
candidate
for
coding
right
now,
if
there's
anything
I
can
help
with.
Certainly
I
can
do
that
more
so
you
know
then,
so
just
let
me
know:
okay,
awesome,
yeah.
E
E
How
do
we
handle
assignments
on
the
triage
issues?
I
noticed
they
are
typically
assigned
to
somebody.
Is
that
the
triage
person
person,
or
we
sort
of
rotating
that
as
well
or
how
do
we
handle
assignments
or
what
was
the
idea
behind
if
there
is
any.
C
D
Yeah,
I
don't
know
like,
I
think
it's
a
little
bit
up
to
preference.
I
think
the
main
the
main
question
here
is
sort
of
keeping
track
of.
Will
we
work
on
this?
You
know,
because
this
is
something
we
need
to
do
or
do
we
need
more
work
and
just
keeping
a
record
in
the
issue.
If
there's
any
specific
actions
to
be
taken,
I
don't
think
things
need
to
be
assigned.
A
But
I
was
under
the
same
impression
that
that
pat
was
just
describing
like
if
you've
spent
any
time
and
have
some
background
on
it
to
assign
it
to
yourself
so
that
someone
else
isn't
starting
from
zero.
Maybe
they
could
just
ask
okay.
What
have
you
done
on
this
as
they
take
it
over
so
yeah?
I
would
recommend
assigning
if
you
spend
any
time
on
it
is
there
are
any
issues
in
here
we
should
discuss,
doesn't
look
like
it.
D
D
I
think
there
are
two
different
ways
of
doing
epic
trees
for
github
that
I
can
think
of
the
one
is
for
jio,
for
example,
because
we're
sort
of
product
facing
we
are
interested
in
increasing
maturity
right.
So
I
group
things
sort
of
by
this
is
all
the
stuff
that
we
think
we
need
to
do
in
order
to
reach
complete
maturity
for
the
database
team
or
the
memory
team.
That's
maybe
not
exactly
the
same,
at
least
not
now,
because
you
know
there's
no
majority
to
reach
right.
It's
like
it's
just
a
you
know.
B
D
You're
interested
in
pushing
for,
and
so
I
think
this
is
maybe
something
sort
of
for
the
longer
term
to
think
about
in
in
databases
like
how
can
we
sort
of
group
areas
that
we
work
in
right
to
make
it
easier
to
understand
how
they
how
they
connect,
and
so,
if
you,
for
example,
click
on
memory
efficiency
here
you
can.
Actually,
if
you
go
back
in
the
previous
screen,
you
can
click
on
the
little
arrow
and
it
unfolds
it.
D
And
then
you
see
all
of
the
like
the
items
that
have
to
do
in
in
memory
with
sort
of
you
know
optimizing
how
memory
efficiency
improvements
can
be
made
right.
If
you
look
at
running
git
web
in
a
memory
constrained
environment,
then
sub
epics
again,
and
so
some
of
these
larger
things
are
more
themes
or
sort
of
meta
groupings.
D
They
would
likely
never
close
right,
they're
more
like
buckets,
but
I
think
it's
still
nice
to
have
sort
of
maybe
like
one
large
meta
epic,
to
hold
it
all
together
some
themes,
but
under
that
you
know
to
have
really
focused
epics
that
are
closable.
I
can
say
like
this
is
a
thing
that
we're
going
to
deliver
right
and
it's
actually
going
to
close,
and
then
it
helps,
I
think,
to
also
track
progress
and
say
like
these
are
all
the
tooling
improvements
we
made
over
the
last
year
right.
D
These
are
all
the
memory
efficiency
improvements
we
made
and
that
helps
at
least
me
keep
track
of
an
ever
growing
list
of
epics
right,
because
then
it
kind
of
pushes
out
the
problem
right.
It's
like
you,
take
issues
and
then
you
make
them
into
epics,
and
then
you
have
an
epic
backlog.
That
is
not
organized
anymore
right,
and
so
this
is
a
way
of
handling
that
just
as
a
highlight
not
as
like
a
highlight,
but
more
like
a.
This
is
a
thing
we
may.
We
may
want
to
do
that
as
well.
At
some
point,.
A
It
I
don't
know
if
anybody
or
everybody's
already
seen
this,
but
I
figured
I'd,
bring
it
up
again.
So
jerry
has
put
together
his
ideas
for
the
database
scalability
working
group.
A
Let
you
all
read
the
merge,
request
and
comment
on
it.
It's
it's
not
focused
on
sharding,
it
does
call
out
sharding
and
the
wording
is
very
generic
in
there,
and
I
don't
know
if
sharding
will
come
back
into
the
discussion
and
by
sharding
I
mean
the
specific
charting.
I
know
there's
a
few
definitions
of
it.
Situs
was
actually
mentioned
again
in
the
discussions
as
a
possible
future
solution.
A
If
the
efficiency
is
there,
even
though
it
costs
us
money,
we
should
reconsider
it
so
right
now
all
things
are
on
the
table
I'll.
Let
you
all
read
jerry's
mr
and
comment
accordingly,
but
this
is
not
specifically
focused
at
the
database
tier
anymore.
There's
a
lot
more
architectural
changes
proposed
in
this.
That
will
largely
impact
the
memory
group
and
some
blueprints
that
are
already
underway,
but
we
will
certainly
be
involved
once
this
starts.
A
A
E
Folks,
just
in
terms
of
priorities
or
focus
this
week,
what
I
would
like
to
do
is
bring
the
testing
in
a
state
that
we
can
handle
two
maintainers.
Basically,
and
once
that
is
done,
I
will
blow
switch
gears
and
go
back
to
where
you're,
going
together
with
the
primary
key
stuff
going
on.
A
Cool
no,
it
sounds
good.
What's
remaining
on
being
able
to
hand
migration
testing
to
the
database,
maintainers.
E
It's
basically
adding
two
more
things
to
query
statistics
and
they
create
log
to
make
this
really
useful,
I
think
and
then
sort
of,
and
then
I
think
we
can
we
can.
We
can
let
maintainers
use
that
and
wait
for
some.
A
Feedback
that'd
be
awesome.
Do
you
think
that'll
happen
this
week
next
week?
How
long
do
you
think
it'll
take
just
curious.
A
E
Week,
wednesday,
okay.
A
C
C
On
top
of
that,
nothing
specific,
though
my
plan
was
to
kind
of
get
everything
gathered
up,
that
I've
been
working
on
and
pushing
up
an
mr,
at
least
so
that
everyone
can
see
what
I've
done
so
far,
and
so
I'm
going
to
be
trying
to
get
that
in
the
state
as
good
as
I
can,
the
next
day
or
two
to
be
able
to,
you
know
push
it
up,
then
we
can
see
what
everyone.