MPIfR Correlator Report (November 2006)
_______________________________________


MPIfR/BKG  MKIV  CORRELATOR  
===========================     

GENERAL

   Current correlator capabilities are listed at:
        http://www.mpifr-bonn.mpg.de/EVN/MK4CORstatus
  

CORRELATOR HARDWARE

   > Present playback status: 
     8 Mark 5As;  (4 to 6 of these can be upgraded to 5B)
     4 Mark 5Bs (1 operational 12/06, rest ~2/07)
     2 Mark 5A I/O-boards ordered as spares
     1 additional Mark 5 unit for Internet transfers
     1 server with 4 TB for Internet transfers

   > The number of functioning station units is down to 8 due to the
     same problems as seen by JIVE (badly manufactured boards).
     Medicina has been asked to repair 2 SUIMs. Haystack has sent 2
     spare SUIMs.

   > In order to be able to upgrade to 12 stations the so-called input
     boards are being repaired/overhauled at Haystack.

   > Modern mainboards have been ordered for all Mark 5 units.

   > For modernising the correlator control computers (HP-UX machines)
     Linux servers and a 14TB Raid have been procured in November. It
     is expected that the old HP correlator computers will be
     decommissioned in 2007.

   > For eVLBI a P2P (point to point) connection at 10 Gb from the
     institute to Dwingeloo to be shared with Lofar will probably be
     realised in January/February 2007. e-transfers are currently done
     via the VIOLA 10 Gb testbed which has a 1 Gb connection to Geant.



CORRELATION SOFTWARE

   > Delivery of Linux correlator software from Haystack is expected
     in December.

   > New Haystack eVLBI software EGAE has been installed and is being
     used for VLBI data transfers.


OPERATIONS

   > Some observations with thin and thick tape are still being
     correlated. They are geodetic observations including Antarctica.
     IVS will stop any tape recording at the end of 2006.

   > e-transfer of VLBI data was successfully done for a geodetic
     EUROPE observation (~500 GB) between Onsala and Bonn using
     Haystack's EGAE software. While a few days before the transfer
     800 Mbps were achieved, before, during and after the transfer
     only about 100 Mbps were possible. The reasons for this lowish
     rate are unclear and are being investigated. The data was copied
     in a scan-based way, first a single scan to a local disk at
     Onsala, then transfered with bbftp to a disk file at Bonn, then
     the next scan and so on. In a 2nd run all the files at Bonn were
     transfered automatically to an 8-pack for further correlation.

   > The transfer of K5-recorded data from Tsukuba will be tested
     soon. Medicina and Wettzell will follow.

   > Production correlation with 1 Mark 5B playback is expected to
     start January 2007. Westford and Parkes will be Mark 5B stations
     real soon.


OTHER

   > Effelsberg will be equipped with one Mark 5A and one Mark 5B
     unit, both will have gigabit Ethernet. It is hoped that the Mark
     5B can be connected to the MK IV terminal with available VLBA
     samplers and a VSIC board. This would allow to record Mark 5B
     format for tests at the JIVE correlator.

   > MPIfR has agreed to take part in joint exploratory VLBI
     observations with Haystack at 4Gbit/s in spring of 2007. 2 Mark
     5B+ units will be provided for Pico Veleta and a Mark 5B unit on
     Plateau de Bure.  About 100 TB of disks were procured for the
     test.

   > High-speed serial-links have been build and sent to JIVE for
     their correlator upgrade.

   > The Swinburne software correlator has been installed on a small
     (~20 machines) cluster. Verification against the MKIV correlator
     with a geodetic data-set is ongoing. It is planned to continue
     the collaboration with Swinburne. An output path to the Haystack
     data format for geodetic applications and phase-cal extraction
     could be implemented in the framework of a PhD project located at
     Swinburne. MPI together with the German geodesists will implement
     data correlation directly from Mark 5 units and will streamline
     the operation so that the efficiency of software correlation will
     become comparable to that of the MK IV correlator. For 2007
     procurement of a much bigger cluster for correlation is
     envisaged.

   > It is expected that the fibre connection between Effelsberg and
     the institute will be ready late summer of next year.

   > After the success of the A/D converter developed at MPI for the
     first generation of EVN-DBBCs, a 2nd generation A/D converter is
     being developed now for the "Virtex 4" DBBC. It is expected to be
     ready early 2007. A diploma thesis project for adding an ethernet
     port is planned for 2007.


W.Alef