From: <stingay_at_email.protected>
Date: Thu, 30 Mar 2006 12:34:42 +1100 (EST)

Guys,

Here is the plan for disks for the May session. Please give me any
comments you have. In particular, comments on the disk space required at
Hobart, Ceduna, and Tidbinbilla. Please note the need to ship around the
sets of 4 x 400 GB disks between ATNF, Ceduna, and Hobart.

Cheers, Steven

For reference, the data rates for v188a and v190a are as follows:

=======================================================================

V188a (Lenc)
-----------------
At, Mp, Pa : 512 Mbps for 12h
Ho : 256 Mbps for 12h
Ti : 256 Mbps for 8h
Note: Higher data rates would be preferred by PI if possible

V190a (Deller)
----------------------
At, Mp, Pa : 512 Mbps for 24h
Ho : 256 Mbps for 24h
Cd : 256 Mbps for 12h
Notes: Higher data rates would be preferred by PI if possible.
There's a possibility that the PI might not require Cd. Need to check!

========================================================================

I've been through and calculated the disk requirements for the May
session. There are two experiments, v188a and v190a (v190a is going to be
split into two LST ranges, so I assume they will be called v190a1 and
v190a2 below). Let me know if you see any problems with the following,
especially the detailed allocation of disks. Perhaps we can discuss this
during the telecon tomorrow.

                        v188a v190a1 v190a2

ATCA 2.8 TB 2.1 TB 3.5 TB
MOPRA 2.8 TB 2.1 TB 3.5 TB
PARKES 2.8 TB 2.1 TB 3.5 TB
HOBART 1.4 TB 1.1 TB 1.8 TB
TIDBINBILLA 1.0 TB ------ ------
CEDUNA ------ 1.1 TB ------

TOTALS 10.8 TB 8.5 TB 12.3 TB

GRAND TOTAL 31.6 TB

Here is a more specific allocation of disks:

v188a
=====
ATCA 7 x 500 GB xraid (second half of March Mopra disks)
MOPRA 7 x 500 GB xraid (half of spare ATNF set of 14)
PARKES 7 x 500 GB xraid (other half of spare ATNF set of 14)
HOBART 1.6 TB removable
TIDBINBILLA 1.6 TB removable

N.B. This assumes that we record at ATCA, Mopra, Parkes for ~11.5 hrs,
rather than the scheduled 12 hrs. Given the Parkes elevation limit, this
is ok.

v190a1
======
ATCA 14 x 180 GB xraid or better (from Swinburne)
MOPRA 14 x 180 GB xraid or better (from Swinburne)
PARKES 14 x 180 GB xraid or better (from Swinburne)
HOBART 1.6 TB removable (spare ATNF 4 x 400 GB)
CEDUNA 1.6 TB removable

N.B. Disk space will be saved by deleting files corresponding to slew
times.

v190a2
======
ATCA 14 x 250 GB xraid (from Swinburne) + 1.6 TB removable
MOPRA 14 x 250 GB xraid (from Swinburne) + 1.6 TB removable
PARKES 14 x 250 GB xraid (from Swinburne) + 1.6 TB removable
HOBART 7 x 500 GB xraid (from Hobart: correlated from March)

N.B. Disk space will be saved by deleting files corresponding to slew
times.

Therefore, v188a should already be covered with the currently available
ATNF 500 GB xraid disks, plus removables at Tidbinbilla and Hobart.

At Swinburne we will have to come up with 42 x 250 GB xraid disks (8.1 TB
under raid5) and 42 x 180 GB xraid disks (5.8 TB under raid5) before the
session, a total of approximately 14 TB. We will do this by releasing
approximately 10 TB of disks that have already been correlated and
verified, correlating data from the March session, and making available
new xraid disks from our stock. We will also have to correlate the March
data from Hobart, in order to release their xraid disks.

Furthermore, following the session, we will have to find another
approximately 11.2 TB of xraid space to copy the data from the removables
to the cluster. We will need to make available new disks from our stock
following the session.

The correlation of the v188a data will be quick. This means that we may
not have to copy the majority of the removables over the Swinburne
immediately after the session. It could even wait until the v188a data
have been correlated.

v190a data will probably take longer, due to mucking about with pulsar
gating, possible multiple phase centres etc.

Any comments?

Cheers, Steven

##=====================================================##
Dr Steven Tingay stingay_at_astro.<!--nospam-->swin.edu.au
Swinburne SKA Project Leader
Associate Professor

Centre for Astrophysics and Supercomputing
Swinburne University of Technology
Mail No H39
P.O. Box 218, Hawthorn, Vic. 3122, Australia

ph: +61 3 9214 8758
fax: +61 3 9214 8797

http://astronomy.swin.edu.au/ska
##=====================================================##
Received on 2006-03-30 11:34:57