Experimental Physics and
Industrial Control System

Luedeke Andreas <[email protected]> · Tue, 02 Nov 2010 11:17:54 +0100

[email protected] wrote on 02/11/10 10:26:
> [...] I definitely have a frequent requirement to get a coherent set of data from a server. Up till now, I have monitored everything and then taken a snapshot whenever I got a monitor from the last item in a processing chain [...] How else can I do this?
One "simple" solution is to combine your data into a single waveform using e.g.
the asub record.
But even then, I'm not sure if your waveform is save from containing "half
new/half old" data:
there is no "standard" mutex semaphore to prevent your data from being
overwritten while it is read.

If your data is distributed over many IOCs, things get more complicated (but you
did know that ;-)
Pulsed machines, like LCLS or the SwissFEL, have very tough requirements on
synchronous data taking over the network.
Those solutions are generic and could be used on a single IOC, too. See e.g.:
http://www.slac.stanford.edu/grp/lcls/controls/global/sw/slc_ioc/bsac/BSACDesign.htm
or
http://cern.ch/AccelConf/icalepcs2009/papers/wep026.pdf
But this is really not simple anymore.

If anyone comes up with a simpler but still reliable solution, I would be glad
to hear it :-)

Andreas

-- 
For every complex problem, there is a solution that is simple, neat, and wrong.
                -- H. L. Mencken

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System