Online telecon notes for May 29, 2002.
Attending:
Discussed pseudo-AEM progress. Difficulties with the data-taking aspect of the module are slowing progress somewhat. Curt didn’t have a time estimate for completion. Byron offered to help with upgrading the COMM card firmware from VHDL files if Goddard doesn’t have the facilities to do it. [It turns out from post meeting discussions that there are other modifications to be made to the ACD’s COMM card to give it pseudo-AEM functionality. The card will therefore need to be returned to SLAC for the upgrade.]
Connie has received the Online CDs to upgrade the ACD’s first test-stand workstation to the most recent release of the Online system. She is currently working on some IP configuration issues. She’ll coordinate with Selim about modifications needed in the system’s hosts file. The other two systems the ACD has are already up to date, so the CDs shouldn’t be applied to them.
Selim gave an overview of what users will need to do to access the SCL system running in VxWorks from LabVIEW screens they create. This involves adding decommutation records to the database, which are then triggered using the pktgen “tlm <dbId> <dcomRecId>” command from either LabVIEW or an SCL script. Selim expects to have a system update available with these features in two weeks time.
Byron indicated that CAL has a legacy Python-based system that he is integrating with the SCL test-stand system. CAL wants to keep the Python-based system alive and use it in addition to the SCL-based system. The Python-based system is already capable of making real-time plots of data collected from a COMM card. Dan has written a socket-based data server layered on Curt’s event data interface. Online is interested in both of these functions for possible integration into a future release of the SCL based system.
There was a discussion on the visualization tool that Online will provide. Under consideration are Hippo, Root and IDL. The original plan to write data files on the NFS partition accessible by VxWorks, convert the files to FITS format on NT and plotting them with one of the packages under consideration is probably not sufficient for all cases.
Ric brought up TKR’s and ELX’s request for external hardware SCL support for electronics testing purposes. Standards under consideration are GPIB and digital and analog I/O on Industry Packs daughter cards. Online requests subsystems with similar needs to get in touch with Ric so that we can work towards a common solution. We’d like to keep the list of such requirements small in order to minimize implementation effort.
Selim described the ODBC connectivity that SCL offers. He and Alicia have tested this. The plan is to use MS Access as a configuration database for tests and to hold test results for report generation. Byron indicated that he is working on an XML based solution for holding configuration information and points out that this is not necessarily out of line with the MS Access plan.
Connie indicated that she couldn’t make much more progress without having the pseudo-AEM hardware. Ric requested that she look into putting together an aemRegress.scl similar to the supplied temCalRegress.scl. This can be done and tested without having the AEM hardware and with the current system. Connie agreed.
We discussed the Online Web page (http://www-glast.slac.stanford.edu/LAT/INT/ONLINE/Default.htm), with access to the discussion forum for posting problem reports, etc.
Ric requested that subsystems submit completed scripts to the Online group for version control and use at SLAC. In particular, scripts setting the hardware (e.g. like the supplied temInit.scl and aemInit.scl) are very useful for those who are not familiar with a particular front end. Byron indicated he would supply us with such an initialization script.
Selim indicated that he would not be sending out system updates in the form of Ghost CDs anymore because of the installation effort required. Future updates will probably be downloadable from the Online Web site, with the occasional Ghost CDs to serve as backup in case of catastrophic failure. These backup CDs would cause non-Online system modifications to be lost if they were used.
Byron inquired about how to avoid having to reboot VxWorks and the rest of the SCL system when TEM power is cycled. Ric said he would look into this since last he heard it required some clean up in the COMM card firmware to make it less sensitive to garbage coming from the TEM during the power up phase. After the meeting Curt reminded him that a partial fix is available by calling some functions that are already available in the latest system release. Ric e-mailed the instructions to Byron and will get them posted in the Online forum. This functionality will appear in the next release through the GLAT and GGLT records.