9.9: Issues and Concerns
$Id: concerns.html,v 1.6 1996/03/14 22:15:43 de Exp de $
The unresolved issues or concerns raised by the information management
system specifications are not primarily software issues. Everything
that has been specified can be or has been done today; most of the design
problems have already been solved. There are therefore no burning
software issues or concerns; feasibility is well-established.
The only "dangerous" hardware assumption made in these specifications is
the emergence of a 10x (6GB) CDROM format in time for the backup and archiving
of DEIMOS image data (Summer 1998). Given current trends in the storage
media industry this assumption is not unreasonable, but could still prove
false. Use of a lower density medium would increase the requirement for
jukebox space; however, this need does not become critical until at least
2 years (expiry of proprietary period for image data) after observing
begins. Backups could be done on lower-density media (at a higher materials
cost and with additional user intervention or small jukeboxes); only
the master image library concept is challenged by the failure of
6GB or better CDROMS to emerge as predicted.
Resources and Policy
It is in this area that we encounter real uncertainties, and the risk of
failure to fulfil our stated requirements and specifications. The public
archive server is the locus of maximum uncertainty.
- Will the observing community respond well to the public archiving of
images? Will there be political conflict over the publication of
images? Will this conflict be severe enough to reduce or eliminate
the usefulness of the archive?
- Will UCO/Lick or Keck provide the necessary on-going staff
("librarian") support to maintain the public archive? If they
do not, will anyone? Will it remain online for only a year or
so before vanishing due to lack of staffing support?
- Will Keck staff be prepared, qualified, and willing to assume
responsibility for the IM system when Lick hands off the DEIMOS
instrument to Keck?
- Given that the public archive does not begin to realize its potential
until circa 2 years after first observing with DEIMOS, will the project
budget address the need for a large jukebox, etc? Will the archive
project be without funding for hardware when the data are finally
publishable? Will we end up with a large library of CDROM and
no way to serve them to the public?
- Will the necessary degree of cooperation from Keck-II telescope
support staff (on MK) be obtainable, to keep the online database current
with regard to physical maintenance procedures? How inaccurate can
this information become, or how much of it can be missing, before
the purpose of the database is defeated?
- Can we integrate an archiving facility with whatever local
data backup system Keck-II telescope will use? What technical or
political roadblocks might we encounter? Will we have enough network
and SCSI bandwidth to do it at all?
- How will the observer logbook facility, annotation facility, etc.
interact with whatever trouble-logging system Keck-II telescope will
be using? will we be able to integrate them at all?
- Will the management and maintenance of the two database servers
and the Web server be adequately coordinated?
University of California