[Date Prev][Date Next][Date Index]

Re: data flow

Bonjour!

Thanks to David for the comments and the new diagram.

There is in fact, between the two flows, only one difference:
At each step where data are sent somewhere, there is a duplicated 
sending to the archive. What happens by having CADC taking care of
the book-keeping is eliminating the need to send the data at two
different places and allowing a centralized book-keeping.

>I propose that the book-keeping be done at CADC because we have many years
>experience at verification of CFHT data as it comes off the telescope. But
>it could be somewhere else. But this diagram is far, far simpler than the
>previous one. 

If the book-keeping is done somewhere else, then we are back to the 
initial data flow, assuming that the CFHT archive should anyway get the 
data from each step of the data flow. 

We have to keep a flow to the CFH community at the various stages 
of the data processing, which means adding a few more arrows to 
David's diagram. That's important as this kind of diagram will be seen
by the community!

So, I attached a new diagram where CADC is playing the book-keeper
in addition to its role as the data distribution/archive center.
I also added a link to CADC from the real time part (data_flow_2).
I also attached a new version of the previous data flow, only identifying
with the same color the book-keeping nodes which should be done at the 
same place, but in case that place is not the archive place (data_flow_1).

Keep the discussion rolling!

Aloha

Christian


-- 
************************************************************
Dr. Christian Veillet,       CFHT Senior Resident Astronomer
Phone: (808) 885-3161   http://www.cfht.hawaii.edu/~veillet/
************************************************************

data_flow_2.pdf

data_flow_1.pdf