CORRELATOR REPORT, EVN MkIV DATA PROCESSOR AT JIVE EVN TOG MEETING, December 2006, Noto 22 November 2006 (statistics cover 11 Mar 2006 - 21 Nov 2006) Bob Campbell Huib van Langevelde Arpad Szomoru SCIENCE OPERATIONS The table below summarizes projects correlated, distributed, and released from 11 March to 21 November. The table lists the number of experiments as well as the network hours and correlator hours for both user and test/NME experiments. Here, correlator hours are the network hours multiplied by any multiple correlation passes required (e.g., continuum/line, >16 station, different phase centers, etc.) User Experiments Test & Network Monitoring N Ntwk_hr Corr_hr N Ntwk_hr Corr_hr Correlated 38 489 654 16 86 86 Distributed 37 464 610 16 86 86 Released 40 486 609 12 70 70 The following table summarizes by session the user experiments still in the queue, with an additional column for experiments not yet distributed (entries = remaining to do / total). The actual correlator time is typically between 1.5-2.5 times these estimates, depending on the number of re-dos or other problems. N_to.corr Corr.hrs N_to.dist session 1/2006 0/17 0/292 0/17 Mar/Apr e-VLBI 0/3 0/22 0/3 May e-VLBI/ad-hoc 0/2 0/22 0/2 session 2/2006 0/15 0/299 1/15 Oct e-VLBI 0/2 0/16 0/2 session 3/2006 5/5 47/47 5/5 (projected Corr_hr) To review some landmarks from recent sessions: 1/2006: 10 VLBA stations recording onto disk during globals; 1st 16-disk experiment 2/2006: 1st all-disk session (no tape involved in any experiment) General: 2 ToO experiments tacked onto the end of session 1/2006 (5-6 March) and another prior to session 2/2006 (15 May) 5 successful open-call e-VLBI experiments to date successful correlation of 3rd MERLIN antenna as a separate station There was a fringe test to Evpatoria on 3-4 August, using the DBBC back-end, with the data recorded onto PC-EVN and subsequently translated to Mk4 format and put onto a Mark 5 pack at JIVE. An ftp fringe test during the observations provided early confirmation that things were going well. In the main correlator, some issues with the DBBCs were noted, but good fringes were seen in half the subbands on baselines to Evpatoria, in polarizations consistent with the LCP that was recorded into both polarization channels. A fringe test to Irbene took place on 13-14 November, at a frequency of 12GHz. At the time of writing, we still don't have all the data in house; very preliminary results suggest similar DBBC behavior but no methanol maser emission has been seen in stations' autocorrelation spectra yet. The Field System has been enhanced to allow full automation of the ftp fringe tests. As in previous sessions, scans indicated as 'ftp scans' in the schedule will automatically be copied from the Mark 5 to a linux file. The new FS feature allows for the chosen scan to be also automatically ftp-ed to the software correlator computer at JIVE. A script running there detects the arrival of new data and automatically performs the correlation and posts the results to a web page which is available to all the stations. The automatic ftp feature will be tried out at a couple of stations during session 3/2006. Prompted by Jodrell, we conducted a test in which they recorded both Cambridge and Darnhall onto a single pack, in a recording mode in which the microwave link from the MERLIN out-stations left "unused" subbands. We copied this pack and were able to correlate Darnhall as a separate station by hacking the correlator-control file appropriately, and fringes were seen in the subbands in which there was actual signals. This test opens the possibility, in certain recording modes, of having three baselines in common between the EVN and MERLIN correlations, increasing the reliability of joining the two data sets together. There should be no additional proposing or scheduling burden on the PI, but at correlation the presence of another station may have consequences for the maximum number of frequency points, and we would need to have sufficient free disk-packs on-hand to which to copy the data. TECHNICAL DEVELOPMENTS We currently have 16 Mark 5A units and 9 DPUs attached to station units (SU) for operations. All of the Mark 5A units are housed inside temperature- controlled cabinets, and are fully connected to their SU, reconnecting a DPU would require swapping 2 cables. Session 2/2006 was our first entirely all-disk session. Developments in e-VLBI are now treated in a separate report to the TOG. New motherboards for the Mark5 units at JIVE are currently under consideration, as is the purchase of a switching router to manage the 16x 1 Gbps lightpaths and the 10 Gbps IP-switched lambda that SURFNet will provide. Preliminary tests of the Mark5A+ firmware upgrade, which enables playback of Mark5B recordings on Mark5A systems, were successful, although further tests will be needed. An upgrade to Mark5B at JIVE, which will allow us to phase out the failing Station Units, is being investigated. The replacement of the old HP correlator control computer by two Solaris servers with dual AMD processors was completed recently and has led to some dramatic improvements, most notably a reduction in (re-)start-up time of the correlator system. Porting the system to a new platform provided the opportunity for a general overhaul and tightening up of the code, and has led to an improved reliability and robustness of the system. Associated with this change of CCC, we have incorporated an updated a priori model whose greatest immediate advantage is output of UVWs more consistent with the residual phases, and offers the ability to input surface meteorology to compute a more realistic dry tropospheric delay contribution as well as VLBA axis tilts. The ParselTongue software, which has been developed in the context of ALBUS, is being used by a fair number of outside users. Many EVN astronomers have employed it successfully for their large or complex data reduction problems. The enhancement of the correlator product with model accountability and phase cal detection data has been slow, due to the lack of manpower. An ALBUS position at JIVE has proven hard to fill. Progress on ionospheric calibration and wide field imaging has been made through a joint effort to process P band wide field imaging data. A method was developed to apply differential ionospheric calibration and a number of wide field imaging issues surfaced which will be addressed in the course of the ALBUS project. A four-station, multiple core top-end computing system was installed for the ALBUS parallel computing project. This system will be used to benchmark wide field imaging, but is also available for visitors. The team that works on distributed correlation has been strengthened. The wide band correlator core now runs on cluster computers and various design decisions have been made on how to port this to Grid computing. New modules were developed to extract the data for the correlator from Mark 5 systems or disk files. A lot was learned from the so-called 7 month demo, in which fringes from the PC-EVN system were obtained. With more people joining in December, the priority will be to use the FABRIC software correlator for ftp fringe checks. USER SUPPORT The EVN Archive at JIVE continues to provide web access to the station feedback, standard plots, pipeline results, and FITS files. Very few PIs prefer to have their data distributed on physical media (DAT, DVD) any more. Public access to source-specific information is governed by the EVN Archive Policy: the complete raw FITS files and pipeline results for sources identified by the PI as "private" have a one-year proprietary period, starting from distribution of the last experiment resulting from a proposal. PIs can access proprietary data via a password they arrange with JIVE. PIs receive a one-month warning prior to the expiration of their proprietary period. In addition to the existing tools for searching the Archive via the FITS-finder utility and the catalogue of EVN observations via the Bologna database, new functionality to connect to the Aladin Sky Atlas and the Sloan Digital Sky Survey have been developed. Significant work has gone into creating support for EVN/Global VLBI proposals in the NorthStar web-based proposal tool. SCHED catalogues have been updated so that all EVN stations, and more recently VLBA ones, have disks as their default recording medium. We continue to contact all PIs once the block schedule is made public to ensure they know how to obtain help with their scheduling, and to check over schedules posted to VLBEER prior to stations downloading them. Following a discussion in the TOG meeting, new safety features have been incorporated into the pre-observation logistics to avoid the risk that stations may observe an older version of a schedule in case PIs change them.