ACNET DAQ Data Recovery
Refresher on How ACNET DAQ Works
One job for each IRM is started behind the AD firewall.
This job pushed data to damen at a predetermined rate.
A cgi on damen catches the data, checks to see how much, if any, is new, and writes one file for each new datum.
The file name contains a timestamp for the data.
Another process continually appends the data onto a larger file, tagged with the run number, and deletes the individual data files.
In order to remerge, resort, repack, etc. the data, the entire file is unpacked.
Each datum is written to it's own file.
The data is then repacked.
How To Do It
- Copy the file you wish to unpack to /acnet/acnet-test/work/irm.
- Go to /acnet/acnet-test/bin.
- Unpack the data by running the 'truncate' program.
Type truncate.irm nnnn m 0 0 where nnnn is the run number and m is the subrun.
The two zeros are the time the run ends in seconds and milliseconds.
Because you are telling the program that the run ended at 0.0, it will pop all the data from the file.
This may take a while.
- Check that the data file is empty.
If it isn't, something is wrong.
- For each file you wish to recover, type manualconcatenate.irm nnnnnnn mmmm iiiiiiiiii jjj kkkkkkkkkk hhhwhere:
This will create a file named nnnnnnn.mmmm.irm in /acnet/acnet-test/work/irm/ directory.
- nnnnnnn is the run number
- mmmm is the sub run
- iiiiiiiiii is the start time in seconds
- jjj is the start time in milliseconds
- kkkkkkkkkk is the stop time in seconds
- hhh is the stop time in milliseconds
- Now you need to add the global headers.
For each file, type addglobalheaders.irm nnnnnnn mmmm iiiiiiiiii kkkkkkkkkk where the variables are defined as above.
This will create a file named boone_nnnnnnn_mmmm.irm in the /acnet/acnet-test/data/ directory,
and another named boone_nnnnnnn_mmmm.info in the /acnet/acnet-test/info/ directory.
Last modified: Fri May 28 13:54 CST 2004