Dear
PEB
Member
The
following
is a
proposal
to
establish
a
short
term
group
of
experts
to
define
the
services
that
a
Tier
2
centre
will
require
from
a
Tier
1
centre.
This
is
needed
to
ensure
that
the
MoU
Task
Force
is
fully
informed
of
the
responsibilities
of
Tier
1
centres
and
to
assist
Tier
2
centres,
particularly
those
in
countries
with
no
Tier
1,
in
their
planning.
It
is
also
needed
to
provide
background
information
for
negotiations
between
CERN
and
countries
that
assume
that
CERN
will
provide
such
services.
I
would
like
to
discuss
this
at
the
PEB
next
week.
Regards,
Les
___________________________________________________________________________________________
Tier
1
services
for
Tier
2
centres
The
current
management
and
organisational
model
for
the
LCG
computing
grid
assumes
that
Tier
2
centres
receive
certain
services
from
Tier
1
centres.
These
could
include:
-
storage
and
distribution
of
raw
and
processed
data;
-
the
provision
of
permanent
data
storage
(mass
storage)
services
for
data
generated
at
the
Tier
2
or
lower
Tiers;
-
database
administration
and
management;
-
operation
of
special
servers
and
gateways;
-
distribution,
installation and
maintenance
of
special
systems
software
(e.g.
grid
middleware);
-
expertise
and
support on
systems
and
networking;
-
support
for
HEP
common
tools
and experiment
applications
software;
-
user
support
services,
such
as
documentation,
tutorials,
end-user
help
desk.
In
some
case
such
services
will
be
provided
by
national
Tier
1
centres,
but
in
many
cases
it
will
be
necessary
for
Tier
2
centres
to
negotiate
this
support
from
Tier
1
centres
in
other
countries.
A
Task
Force
has
been
established
to elaborate
the
LHC
Computing
MoU.
This
will
cover at
least the
Tier
0
and
Tier
1
centres,
and
so
it
is
necessary
now
to begin
to
define the
services
that
Tier
1
centres
must
provide
for
Tier
2s.
At
the
same
time
a
number
of
Tier
2
centres
are developing
their
plans,
which must
include
identifying
and
negotiating
the
services
they
require
from
Tier
1.
A
small
group**
is
at
present
working
on
a
re-assessment
of
the
computing
resources
required
by
the
LHC
experiments
from
2006
onwards
at
Tier
1
and
Tier
2
centres,
to
inform
the
Computing
MoU
Task
Force.
I
would now
like
to appoint a
small
group
of
experts
to produce
a
description
of
the
services
that
will
be
required
by
a
Tier 2
centres
from
a Tier
1,
to
further
inform
the
MoU
Task
Force,
and
as
information
for
people
planning
Tier
2
centres.
The
group
should
start
work
as
soon
as
the
report
on
computing
resources
is
available
-
scheduled
for
mid-April.
The
target
would
be
to
report
by
mid-June.
I
propose
that
this
group
consist
of
four
experts,
two from
Tier 2 centres and
two
from
Tier
1
centres.
These
would
preferably
be
people
that
have
been
studying
the
issue
of Tier
1
-
Tier
2
support,
should
include
someone
from
FZK,
which
has
been
discussing such
support
with
other
countries.
__________________________________________
**
Resource
re-assessment
group:
ALICE
-
Alberto
Masoni,
ATLAS
-
Dario
Barberis,
CMS
-
David
Stickland,
LHCb
-
Nick
Brook,
Fabric
Manager
-
Bernd
Panzer,
Networking
-
David
Foster,
LCG
Resource
Manager
-
Chris
Eck,
chair
-
Les
Robertson.