IPAC 2MASS Working Group Meeting #92 Minutes, 4/09/96

Attendees: R. Beck, T. Chester, R. Cutri, T. Evans, J. Fowler, L. Fullmer, G. Kopan, B. Light, H. McCallon, J. Rosenberg, S. Wheelock

AGENDA

1. Action Items From the PS CDR
2. Project Status Updates
3. Memory Upgrades
4. Inputs to the QUALITY Subsystem

DISCUSSION

1. Action Items From the PS CDR -- Due dates for the Action Items from the Point Source CDR have been renegotiated in a telecon with NASA Headquarters. R. Cutri will distribute a new list with due dates. Cognizant engineers are requested to review these dates and respond to R. Cutri if any problems are apparent. The closure mechanism for action items remains an email memo to R. Cutri with "cc: 2mass" explaining how the item was closed. Subsystem liens can be closed by making the appropriate modifications to the software and documentation and redelivering.

2. Project Status Updates -- R. Cutri reported on several updates to the project status as follows.
   • A contract for the second telescope has been signed.

   • The first telescope is on schedule for completion by December '96, with a possible slip of about a month to have the mirror overcoated with silicon dioxide.

   • J. Huchra is meeting with the Forest Service to clarify the Mt. Hopkins constraints.

   • A snag has developed in the MOU with NOAO involving an indemnification clause for design errors by UMASS.

   • There are now three working engineering camera arrays.

   • The flexure problem reported at the CDR has been traced to the array translation stages, which will be replaced with fixed mounts.

   • The image distortion has been traced to a phenomenon in the electronics; this was found by rotating the arrays and noting that the distortion retained its orientation with respect to the arrays, not the optics.

   • Science-grade arrays will be used for the May-June observing run.

3. Memory Upgrades -- Since the new hardware configuration involves 8 ultrasparc CPUs for the production cluster instead of 24 Sparc/1000 CPUs, some concern has developed over the amount of memory per CPU. This was resolved for the production cluster in this discussion, which concluded that doubling the memory per CPU was prudent and not very expensive, so that the configuration is now seen as two machines with 4 CPUs each and 512 MBytes of high speed memory each.

   It was not clear whether similar upgrades were needed for the analysis and data base machines. T. Evans agreed to review the needs with B. Narron and issue a report.

4. Inputs to the QUALITY Subsystem -- Since the system integration is drawing nearer, it is time to organize the system wide information flow into the QUALITY subsystem. It is too early to request SIS level definitions of the information from each subsystem, and in fact very detailed definitions would be a bit premature at this time, but a prioritized list of possible types of information should be formulated by each cognizant engineer. These lists will be discussed at a working group meeting in a week or two. The guidelines for development of the interfaces are as follows.
   • Table file format should be used unless clearly inappropriate.

   • The great majority of analysis should be done in the generating subsystem, with a relatively reduced number of parameters going to the QUALITY subsystem, rather than having large dumps of possibly interesting parameters flowing through the system. The latter should be confined to the subsystems' own propietary file set.

   • Limits on certain parameters, as well as other types of information, will need to be encoded in the NAMELIST input to QUALITY, implying an usual amount of involvement in that subsystem by cognizant engineers of other subsystems.

   ---

   • 2MASS Project
   • IPAC 2MASS page
   • IPAC Home Page