This website is no longer maintained. Its content may be obsolete. Please visit http://home.cern for current CERN information.
I have made a file containing sample post-mortem data and written an associated BinX description.
In the directory /user/poccdev/projects/fgc/sw/clients/pm/ you will find:
pmdata.bin Raw binary data
pmdata.txt Decoded binary data in ascii
pmdata.xml BinX description of binary data
I think that the BinX should be correct, however I haven't validated it in any way. If there are any errors, typos or problems, please let me know.
The data contains 3000 samples (1 minute's worth) which represents a single gateway's worth of post-mortem. Up to 30 power converters can be attached to a single gateway. In the case of the sample data, some of the function generators are not powered, however the size of the data remains the same as for a fully populated bus. A bit in the status byte for each channel indicates whether the data for that channel is valid or not. This can be ignored during these initial tests, but it is something we will have to tackle later on. Channels for which data is not valid are not included in the ascii decode.
It should also be noted that in the ascii file, the gateway's data is referred to as "Channel 0" and the data for the power converters "Channels" 1-30.
The first challenge should be for someone to decode the binary data using the BinX description and try to obtain the values that appear in the ascii file.
Oh and everything is in network byte order (big-endian).
For the first tests of validation of BinX I agree with Steve and these first tests should be finished before the end of February in order to define choices.
After:
* To test CMW under the conditions envisaged by PM
* How to manage PM data in the PMS.
* To test the load response of the networks (QoS approach ...)
* To define a strategy for the creation of the data warehouse which will contain
the PM data (LHC Logging, private ...)
* To define a strategy for the PM clients
* To define analysis tools ?