EVN Technical & Operations Group (TOG) Meeting OAN, Yebes, Spain Monday 12th November 2007 Introduction ------------ The EVN TOG held a meeting on 12th November 2007 at Yebes Observatory, Spain. Approximately 40 people attended the meeting. Alef (Chair) welcomed the participants. De Vicente outlined local arrangements. Reports and presentations from the meeting are available on the web at http://www.mpifr-bonn.mpg.de/div/vlbicor/tog_chair/togreps07/ 1. Approval & last minute additions to Agenda --------------------------------------------- The agenda was approved, with the addition of a discussion of 4Gbps recording to be led by Whitney. 2. Review of Action Items from last meeting ------------------------------------------- The TOG reviewed action items from the previous meeting. 1. Action: Garrington to make the Mk2/Lovell phase referencing technique available as a working process for users - the technique is now available to users. Assistance with scheduling is available from JIVE. Confirmation from the Director is needed before this is widely advertised. Additionally, there are some problems affecting the Lovell telescope sensitivity which are currently under investigation - these problems should also be documented. Action Item: Following approval by the Director, Garrington to produce documentation for the Mk2/Lovell phase referencing technique. 2. Action: All stations should look at data regularly with chchk program, use it to locate significant RFI and report the results to Polatidis and CRAF representatives, and use it to check phase-cal throughout the session - some stations do this, but not much feedback is being provided to the relevant people. The chchk program can be run in gaps, or for example on ftp data files. Alef urged all to monitor RFI and send information. 3. Action: Polatidis, Lindqvist and Burgess to define standards for RFI monitoring and report at the next meeting - not done yet. Action remains open. Persistent RFI should be noted in the EVN status table for the benefit of users, in particular for spectral line experiments. 4. Action: All stations should look at pipeline results available from the JIVE webpage, in particular amplitude corrections found by selfcal on strong, compact calibrators - most stations do. Pipeline results are found to be helpful. It is important that NMEs are pipelined quickly at JIVE to provide feedback well in advance of the next session. JIVE should send email to EVNtech when NME pipeline results are available. NME calibration files should be made available quickly (within 2 weeks after the session). There was some discussion of ANTAB file production. Most of the effort is in producing the rxg files. 5. Action: All stations to provide detailed Wx and RFI comments in experiment logs - Wx is most important for high frequencies (not observed in recent sessions). This is standard for geodetic stations. Wx is now in the logs also for Jodrell, Westerbork and Torun. Format to be checked for Torun. done 6. Action: All stations to continue monitoring Tcal throughout sessions - this can be done by running ANTABFS and plotting the results. Alef requested that all stations check Tcal occasionally. 7. Action: de Vicente to report on implementation of continuous tpi on VLBA systems - not done as the Yebes antenna is not yet ready to test this. Continuous tpi is not yet implemented in the Field System; a special version of the software was developed by Graham. 8. Action: Stations to produce ANTAB and RXG files as soon as possible after the end of a session, within 2 weeks - this is sometimes fine, sometimes an issue. Especially for e-VLBI it is often important to have ANTAB files as soon as possible - generally rxg files from the previous session can be used. JIVE to inform Alef if there are problems, so Directors can be asked to prioritise calibration if insufficient time is available at stations. 9. Action: All stations to forward FS error log files to Himwich and include details of what the FS was doing when the crash occurred - this is being done. There have not been many crashes to report recently. 10. Action: Reynolds and Walker to incorporate frequency agility information in SCHED - not done yet. This has dropped in priority below working to incorporate Mk5B and C in SCHED. 11. Action: All stations to look at NME reports sent by JIVE - this is being done. 12. Action: Gunn to send email before each session when the final versions of all schedules are ready for download - several schedules last session were submitted or revised after the deadline which complicated the process. 13. Action: Reynolds to report on the feedback re. amplitude calibration methods from spectral line users at JIVE - done. The conclusion is that providing good quality amplitude calibration data cannot be neglected for spectral line experiments. 14. Action: All to check that Mk5 modules are placed squarely on a flat surface when received, otherwise connectors are easily damaged when bent 8-packs get inserted in Mark 5 units - ongoing recommendation. There have been no problems recently. 15. Action: All stations to try to run the FS diagnostic tests and investigate the results - this is done at some stations and is found to be moderately useful. Permanent action items/recommendations for the TOG to be moved into a single summary. 16. Action: Reynolds to work with Rene Vermeulen on the guide for polarization calibration using WSRT local interferometry data - no guide has been produced yet. Zsolt Paragi is now leading this effort, with Tony Foley the WSRT contact. 17. Action: Paragi to implement labelling of fringe plots with sky frequency and sideband - this now needs to be done for the new JIVE software correlator. 18. Action: JIVE webmaster to add URLs where stations can look for experiment feedback to 'VLBI links' on EVN web site - done 19. Action: All stations should condition disk packs if time permits, especially those which are to be used for 1Gbps recording - Alef suggested that label is changed to 512 Mbps if a disk pack is found to be not suitable for 1Gbps recording. 20. Action: All stations which do both astronomy and geodesy should clearly distinguish between astronomical and geodetic 8-packs and ensure there is no "leakage" into the wrong pool - there was some discussion of this issue. Alef noted that disks are still an expensive and scarce resource and mixing pools can create problems for the scheduler and should be avoided. 21. Action: Alef to distribute astronomical/geodetic disk database information to Haystack - not done yet. Owners of mixed pools (Cassaro, Orlati, Quick, Mujunen, Lindqvist) were requested to give information directly to Alan Whitney. It was suggested that the information be put on the TOG wiki created by Tony Foley. Foley to add a wiki page for astro/geodetic disk labelling information. 22. Action: All stations to ship disk packs as soon as they are full, at least once per week - stations are reminded of this, following the Bologna rules. 23. Action: Disk packs should be shipped with one European and one US address on covers so they can be easily returned in case a shipment is lost in transit - done at Bonn and NRAO. JIVE should also follow this procedure. Very few if any diskpacks have been lost in shipment. Campbell noted that no incoming diskpacks have ever been lost. 24. Action: Foley to start a wiki page for FS documentation, which all users can contribute to - done, and email sent by Foley (on 11 December 2006). A namespace for the TOG was added to the dokuwiki at JIVE. NME reports are now placed here as well. The wiki has been not much used yet by others. It should be linked from the TOG web page. 25. Action: Burgess to look into making the EVNtech mail archive accessible on the web - done. The archive is password-protected to prevent email addresses being "harvested" by spammers. 26. Action: All (especially non-geodetic) stations to check FS weather logging, as the information can now be used - see response to action item #5. 27. Action: Campbell to look into producing phase-cal plots at JIVE - phase-cal information is not yet included in the Measurement Set data produced at JIVE. This has dropped in priority below Mk5B implementation. 28. Action: JIVE support scientists to include checking of sampler statistics as part of NME feedback - a checkbox for sampler stats has been added to the NME report. In general sampler stats are checked for every experiment, and any problems reported. 29. Action: JIVE support scientists to look into a scheme to estimate the quality of pipeline amplitude calibration results - support scientists to insert a comment on the pipeline page to indicate the reliability of the amplitude calibration results. Occasionally there may be a problem in a particular experiment producing misleading results, but this does not happen often. Stations should look at all experiments from a particular session to get an idea of the calibration quality. 30. Action: Friends to provide gain curves free of opacity corrections (as if taken outside atmosphere). Campbell to inform users. Also PIs of 22 GHz experiments to be informed that Onsala has a different method of calibration at 22 GHz (hot load instead of noise diode), which means that Tsys values are already corrected for opacity - stations should aim to produce opacity-free gain curves for the next session. This is consistent with what the VLBA provides. Lindqvist and Graham to produce a document describing what stations and users need to do. JIVE support scientists to inform users. There was some discussion of how to measure receiver temperatures needed for APCAL - hot/cold load measurements are often difficult at low frequencies due to the large size of the feedhorn. 31. Action: Campbell to determine an appropriate place to put WSRT and Effelsberg calibrator flux measurements - in progress. The measurements are not on the web yet. Zsolt Paragi communicates WSRT results privately to PIs in some cases (for e-VLBI experiments). 32. Action: Smythe to send email reminding stations of procedures for checking Mark 5 recording - done (on 22 January 2007). 33. Action: Smythe will contact Brisken to organise a teleconference with NRAO on Mark 5 software status - (SDK 7) done. 34. Action: Alef to send email to stations regarding disposal of tapes - this was discussed later. 35. Action: Alef to ask Mantovani if Directors can donate money to purchase spare Mark 5 backplanes - Alef asked if any other spare parts are needed. There was no feedback yet. 36. Action: All to send constraints for the next TOG meeting to Alef - done. 3. Reliability/Performance of the EVN ------------------------------------- Cimò reported on reliability and performance of the EVN. The detailed report is available on the web. The median EVN reliability indicator (ERI*) showed a big improvement with the introduction of disks, from ~0.6 to 0.85. The ERI* for e-VLBI experiments is ~0.6, which includes time lost at the correlator due to restarts being needed, as well as occasional station failures. Thus e-VLBI is already effectively matching the reliability of tape recording, and continued improvements are expected. There was some discussion of Mk5/FS control computer communication problems requiring reboots, and of slow disk problems. Whitney noted that conditioning should reveal problem disks. The FS should also be capable of recognising a slow disk and giving a warning. Action Item: Himwich to be contacted regarding implementing slow disk warnings in the Field System. It was noted that the current problems occurring with some VCs will be solved by the introduction of DBBCs. Alef requested feedback from the last session. Lindqvist requested that NMEs are scheduler earlier in the day (CET) so that ftp fringe checks can be done the same day. Action Item: EVN Scheduler to try to schedule future NMEs in the early part of the day in Europe. Torun had problems with disk2file in the ftp fringe tests due to a Mk5 software bug. The latest version of the code was downloaded from Haystack for e-VLBI tests, but subsequently the bug was found and the software version retracted. The problem has since been fixed, and the issue of version control in the automatic download procedure is recognised. Onsala lost two experiments in the last session due to high winds. Also they had a problem with the new S/X receiver being warm. Noto was affected by a subreflector movement problem, which requires installation of a spare part to fix. The effective gain will be azimuth-dependent, not just elevation-dependent. The gain curves in the rxg files are affected by the problem. Users should be warned. Also a Mk5 problem was encountered, where recording was slow in one experiment. Recording was OK again after a Mk5 reboot. The Chinese stations did not participate in the October sesion due to their lunar project. An Tao noted that in the June session, Shanghai had a problem with variable Tsys at L-band in experiment EG037. Robledo successfully participated in one observing session in 2007. Tests of loading on the antenna structure are ongoing. FS version 9.9.0 was installed to solve the previously encountered FS crash. Orlati commented that notes in the schedule are helpful if the PI wishes the phase-cals to be turned off (there was some confusion over this in one special case last session). Also one experiment (going to Socorro for correlation) used more disk modules at Medicina than expected, due to long recording periods without gaps. There was some discussion of this issue. As the EVN cannot switch disks on the fly, shorter intervals between gaps are desirable. On the other hand, the NRAO correlator cannot synchronise quickly after a gap, so for experiments correlated there, fewer gaps are preferred to avoid losing a lot of data. Action Item: Reynolds to discuss with Craig Walker a change to the current 90-minute no gap warning in SCHED, to reduce the maximum interval for EVN stations. Gunn reported that there were several problems at Cambridge during the last session. The Cambridge formatter lost sync in N07M3, but was reset for the user experiments. Also the receiver carousel was noticed to be in the wrong position during EH022, which required a last-minute software change and resulted in some time being lost. In GP044, there may be sporadic data loss due to pointing offsets, resulting from a rogue telescope control software process to nod off source which had been running for calibration tests. Before the cause of the problem was tracked down, a new schedule was made with corrected offsets manually inserted for each scan. There were two failures at Hartebeesthoek in the last session. The first was due to rain leaking into an electronics rack in the equipment room, resulting in the loss of 1 hour of a 2 hour experiment. The second problem was a swapped cable but no data were lost as a result. Westerbork had a problem sending the ftp fringe test data, due to a local file server problem. The ftp data were sent manually after the NME finished. 4. Amplitude Calibration ------------------------ Cimò reported on the current status of amplitude calibration. The importance of timely calibration data for e-VLBI experiments was emphasised, particularly as rapid analysis is often important for e-VLBI experiments. RXG files from the previous session can be used. Gunn pointed out that e-VLBI sessions may use different receivers than the normal sessions, in which case extra telescope time for calibration and time for analysis may be required. If there are problems getting the necessary time for calibration, Alef can contact the Directors. In general, calibration accuracy at 6cm is good. There are some possible RFI problems at 18cm and 5cm. Ambrosini commented that RFI reports should be official, as this enables action to be taken. Alef also noted that own interference needs to be examined. Gunn reported on the status of Cambridge calibration. Currently the gain cannot be measured across the microwave link. The data are there in principle, but a conversion factor is needed to get the Tsys out. Some tests were done with nodding observations. Currently the data conversion is time consuming as there is no standard software available for this procedure. Also the e-MERLIN upgrade to optical fibre transfer, planned for next year, will make the microwave link obsolete, so the effort to extract Tsys from the Cm data may not be worthwhile. The failure of the cal diode at the Lovell telescope seen in some recent experiments was not seen at the station. The cause of the problem is unknown. In principle the telescope controllers could be asked to monitor the TPI values. Reynolds commented that it may be useful to look at the frequency dependence of calibration accuracy within L-band. There was some discussion of AIPS limits on the number of Tsys entries in ANTAB. The limit in AIPS has been recently increased but could be increased again, as there may potentially be problems for long experiments. In cases where the number of Tsys entries exceeds the limit, they are decimated by commenting out some fraction of the entries. There is no advantage to averaging. Action Item: Orlati to contact Himwich regarding separation of LCP and RCP receiver temperatures in the FS and GNPLT, as they may be different. There was not much discussion of DBBC and calibration (Tuccari was absent), except that 80 Hz continuous calibration will be implemented. Action Item: Uwe Bach, Zsolt Paragi and Tony Foley to sort out how to make Westerbork and Effelsberg fluxes available before the next session. Reynolds reported that automatic flagging is not working properly at Westerbork because of the 10s granularity limit. Flagging to the nearest second would be more useful. The SCHED model could be used to produce the flag file, but very accurate telescope parameters are needed. Action Item: Reynolds to look at Westerbork automatic flagging problem. Borkowski gave a short presentation on the role of DPFU quantity in EVN calibration, adding some discussion also of the gain polynomial normalization. In the current CL experiment procedure it is not possible to determine the DPFU. However, it was noted that temperatures are needed sometimes, e.g. for opacity correction, and therefore it is important to try to get the DPFU approximately correct. Graham also noted that Tsys values are useful for system engineers. Borkowski requested better documentation for GNPLT and ANTABFS. Alef commented that the TOW mini-workshops are a useful source of knowledge. All stations are encouraged to attend TOW meetings. Documents and presentations from the last TOW are available online. Mujunen noted that stations should not try to determine the DPFU from GNPLT fitting, but use an alternative method. Lindqvist added that hot/cold load measurements are ideal but are difficult at most stations. It was noted that while Tcal values may vary, DPFU should not. 5. Mark 5 --------- Whitney reported on Mark 5 status. He noted that little feedback had been received so far on the Mark 5C specs sent around. There was some discussion of disk failure statistics and record keeping. Leeuwinga and Campbell were asked to send JIVE's record of failed disks encountered to Dan Smythe who keeps a record at Haystack. Alef and Campbell reported on EVN Mark 5 status. There is still a lot to test at JIVE, but currently it looks possible to correlate recorded Mark 5B data. e-VLBI is not yet working with Mark 5B. There was some discussion of upgrades to the new Debian version and SDK 7 software. Whitney suggested stations try to first update on a new disk to check that the upgrade works, keeping a copy of the old version in case of problems. Action Item: Whitney to send information on Mark 5 OS and control code package upgrade procedure. Alef presented the disk inventory. The EVN currently has ~1100 TB recording capacity, therefore more than 500 TB should be available for each session, and 1Gbps recording is okay as the default for continuum observations. Alef plans to put the disk inventory on the wiki, so that stations can update the table themselves, and also provide information for JIVE on what disk space they have to hand before sessions. The information in TRACK is not always correct or up-to-date because stations cannot manually correct entries. Alef requested that updates are sent to Bonn where values can be corrected, and "dead" 8-packs and tapes can be deleted. Bertarini reported on the status of tape disposal. In most cases it is OK to dispose of tapes but there are some exceptions. Nobeyama wants to keep some of its tapes (named NO*) so these should not be destroyed. It is not yet clear if they will pay for shipment of their tapes. There are also some problems for tapes which are US government property, such as those with a NASA label or Air Force (AF0). Tapes from Caltech, CSIRO, EVN, Haystack, JPL can be disposed of, as can those with NGS and NLL labels. In some cases the origin of labels was uncertain or unknown. To get rid of tapes, send VSN and check codes to Alef for updating in TRACK, and find someone who wants the tapes for the aluminium. Mark 5 logistics were discussed. Regarding disk replacement, stations can tell JIVE when they want to replace disks so that smaller disk-packs can be sent back. In general JIVE tries to send disk packs 1 month before the beginning of each session, which was deemed to be long enough to condition packs. There was some discussion of criteria for replacing slow disks, which are currently somewhat arbitrary. Later, during discussion of the Field System, it was decided that Dan Smythe should come up with specific criteria for "slow disk" alarms. In general, for 1 Gbps recording it is better to change the pack if disk performance is doubtful. Himwich suggested that it could be possible to implement automatic bank-switching in the case of slow disks at high data rates. JIVE does not see the cost of shipping when the station pays. It was decided to establish a cost database. This would help with finding the cheapest shipping company and provide savings estimates to the Directors. 6. Data acquisition system for 4 Gbit/s, wide IFs discussion ------------------------------------------------------------ More bandwidth was a requirement of science drivers in the EVN2015 vision. NRAO aims to upgrade the VLBA to 4Gbps recording by the end of 2008, and the EVN needs to decide how to stay competitive and enable Global wide-band experiments. After some discussion, a committee was formed with the following people, in order to define a technical solution for 4 Gbps recording in the near-term to propose to the CBD: Alef, Graham, Lindqvist, Mujunen, Ritakari, Szomoru, and Tuccari. Whitney is available for consultation, and maybe Jon Romney. 7. Field System --------------- Graham presented the Field System status report on behalf of Himwich. The presentation is available on the web. 8. SCHED Developments --------------------- Reynolds reported on SCHED developments. There has been no release since the last TOG meeting, but one is expected soon. During a recent mm session there was a problem with byte number limit being exceeded; the byte limit will be increased in the new release. Tape support will not be removed yet. There will be some support for Mark 5B modes. The Mark 5B recording modes are different from Mark 5A modes currently enforced by SCHED; additional modes therefore need to be allowed for Mark 5B. There was some discussion of the importance of frequency agility being incorporated into SCHED. Campbell estimated that up to ~25% of proposals request frequency switching. There is currently nothing to stop users scheduling different frequencies, but SCHED does not have information about the receiver change times, therefore no warnings will be issued about impossible setups. It could be useful to add frequency agile EVN telescopes to some global experiments. Stations were urged to keep catalogue information up to date and send in any modifications. 9. Digital BBCs --------------- Alef read the presentation on DBBC Status on behalf of Tuccari. It is expected that the DBBC design will be finished in early 2008, therefore DBBCs should be available in late 2008. It is important to find a cheap production option so that DBBCs can be made available to stations at a reasonable cost. Whitney gave a presentation on the status of the DBE at Haystack Observatory. 10. Spare parts --------------- Gunn reported that Burgess is maintaining the spare parts web page. Alef requested more information on this to be sent for the next TOG meeting. 11. e-VLBI status ----------------- Szomoru reported on EVN e-VLBI efforts and EXPReS developments. The full presentation is available online. The idea of adaptive observing, soon to be tested with Westerbork, is that the schedule can be changed on the fly, the FS daemon looks for a new schedule, gets it and DRUDGes it automatically. Individual stations can now be reconnected without a job restart, which saves about 15 minutes each time. General improvements to make the correlator control code faster also improves the efficiency of "normal" correlation. For the SCARIe FABRIC project, JIVE has access to 2 clusters for software correlator testing. Increases in bandwidth will require many more Teraflops than the current MkIV correlator - expected bandwidth increases are faster than increasing PC capability. Stations were requested not to change over to Mark 5B yet if possible, because e-VLBI does not work yet with Mark 5B. There is some enthusiasm to test e-shipping, but some logistical issues still need to be solved. Bertarini reported on e-VLBI status at Bonn. Several additional stations will start testing soon. Alef reported on the Effelsberg fibre link status. Fibres are being patched and tested for the new link for LOFAR and e-VLBI. Whitney reported on the status of e-VLBI at Haystack. Data transfers are done routinely over a 1 Gbps link. A dedicated 10 Gbit Internet2 data connection from Haystack to Washington DC has recently been installed, courtesy of Lincoln Lab. There is some limited support for geodetic e-VLBI, but there is no NSF support for formal e-VLBI projects, as e-VLBI is seen as low priority compared to the EVLA and ALMA. 12. Station & correlator reports -------------------------------- Station and correlator reports were not presented but are available online. 13. AOB ------- Ambrosini gave some recommendations on CRAF matters. He stressed the importance of making official reports if any action is to be taken on RFI. National regulatory bodies are responsible for stopping illegal interference. It is important to set a good example at stations. Both spectrum usage and non-usage of allocated bands needs to be justified. For VLBI, there may be some "acceptable" level at which observations are not significantly affected. Buiter announced that this was his last TOG meeting. He will be retiring on 1 March 2008. He has been involved in VLBI at Dwingeloo since 1979, and his first EVN "Technical Working Group" meeting was in April 1992. He thanked the TOG for many years of successful collaboration and wished good luck for the future. Alef thanked Buiter for his contribution to the TOG over the years. 14. Time and place of the next meeting -------------------------------------- The next meeting will probably take place around 1 April 2008, at a location to be decided. Alef thanked all for a productive and relaxed meeting. List of Action Items -------------------- Action Item 1: Following approval by the Director, Garrington to produce documentation for the Mk2/Lovell phase referencing technique. Action Item 2: Polatidis, Lindqvist and Burgess to define standards for RFI monitoring and report at the next meeting. Action Item 3: Format of Wx information in logs to be checked for Torun. Action Item 4: de Vicente to report on implementation of continuous tpi on VLBA systems. Action Item 5: Reynolds and Walker to incorporate frequency agility information in SCHED. Action Item 6: Paragi and Foley to produce a guide for polarization calibration using WSRT local interferometry data. Action Item 7: JIVE software correlator group to implement labelling of fringe plots with sky frequency and sideband. Action Item 8: Owners of mixed astronomical/geodetic disk pools to give database information to Alan Whitney. Foley to add a wiki page for astronomical/geodetic disk labelling information. Action Item 9: Campbell to look into producing phase-cal plots at JIVE. Action Item 10: Stations to produce opacity-free gain curves for the next session. Lindqvist and Graham to produce a document describing what stations and users need to do. JIVE support scientists to inform users. Action Item 11: Alef to ask Mantovani if Directors can donate money to purchase spare Mark 5 backplanes, and any other spare parts if needed. Action Item 12: Himwich to be contacted regarding implementing slow disk warnings in the Field System. Action Item 13: EVN Scheduler to try to schedule future NMEs in the early part of the day in Europe, to better facilitate rapid response to fringe test feedback. Action Item 14: Reynolds to discuss with Craig Walker a change to the current 90-minute no gap warning in SCHED, to reduce the maximum interval for EVN stations. Action Item 15: Orlati to contact Himwich regarding separation of LCP and RCP receiver temperatures in the FS and GNPLT, as they may be different. Action Item 16: Bach, Zsolt Paragi and Tony Foley to sort out how to make Westerbork and Effelsberg flux densities available before the next session. Action Item 17: Reynolds to look at Westerbork automatic flagging problem. Action Item 18: Whitney to send information on Mark 5 OS and control code package upgrade procedure. Summary of "Permanent" Actions/TOG recommendations -------------------------------------------------- SESSION PREPARATION: Gunn to send email before each session when the final versions of all schedules are ready for download. All should check that Mk5 modules are placed squarely on a flat surface when received, otherwise connectors are easily damaged when bent 8-packs get inserted in Mark 5 units. All stations should condition disk packs if time permits, especially those which are to be used for 1Gbps recording. If a disk pack is found to be not suitable for 1Gbps recording, the label can be changed to 512 Mbps. All stations which do both astronomy and geodesy should clearly distinguish between astronomical and geodetic 8-packs and ensure there is no "leakage" into the wrong pool, as mixing pools can create problems for scheduling. Disk packs should be shipped with one European and one US address on covers so they can be easily returned in case a shipment is lost in transit. DURING SESSIONS: All stations should look at data regularly with chchk program, use it to locate significant RFI and report the results to Polatidis and CRAF representatives, and use it to check phase-cal throughout the session. The chchk program can be run in gaps, or for example on ftp data files. All stations should monitor Tcal throughout sessions. This can be done by running ANTABFS and plotting the results. All stations should forward FS error log files to Himwich in the event of a crash, including details of what the FS was doing when the crash occurred. All stations should try to run the FS diagnostic tests and investigate the results. All stations should ship disk packs as soon as they are full, at least once per week, following the Bologna rules. SESSION FEEDBACK: All stations should look at pipeline results available from the EVN data archive pages at JIVE, in particular amplitude corrections found by selfcal on strong, compact calibrators. JIVE support scientists should include a comment on the quality of amplitude calibration results, especially to indicate cases where a problem may have occurred and the amplitude correction factors are unreliable. NME calibration files should be made available as early as possible. All stations should look at NME reports sent by JIVE. NMEs should be pipelined as early as possible and email sent to EVNtech when the results are available, to ensure feedback is provided well in advance of the next session. Stations must aim to produce ANTAB and RXG files within 2 weeks after the end of a session. For e-VLBI, RXG files from the previous session can often be used. ANTAB files for e-VLBI experiments should be produced as soon as possible as rapid analysis is often a high priority for these experiments. JIVE should inform Alef of any problems, so that Directors can be asked to prioritise calibration if insufficient time is available at stations.