On 10/10/11, John Doe <jdmls@xxxxxxxxx> wrote: > From: Eero Volotinen <eero.volotinen@xxxxxx> > >> 2011/10/10 hadi motamedi <motamedi24@xxxxxxxxx>: >>> I have installed an announcement application on my centos 6.0 server >>> that calls for putting specific voice announcement files under >>> /usr/local/srf/bin/prompt to be played in response to certain >>> conditions occurred . There are a huge number of files in the >>> announcement directory and it seems that just one of these voice files >>> is corrupt . Can you please let me know how can I trace in real time >>> to see which application is going to use this folder and which of >>> these files will be accessed at the moment ? My goal is to find that >>> corrupted voice file in real time . >> >> How about something like this: >> watch -n 1 lsof /path/to/files > > Or maybe: > inotifywait -m -e access --format "%T %f" --timefmt "%D %T" -r > /path/to/files > > JD > _______________________________________________ > CentOS mailing list > CentOS@xxxxxxxxxx > http://lists.centos.org/mailman/listinfo/centos > Excuse me, the announcement application program is accessing this folder from time to time to play the appropriate voice announcement file . As there are a huge number of voice files inside this folder, so I need some way to trace to see which file is being accessed when hearing the corrupted voice file . I tried for your "watch" & "inotifywait" utilities but I didn't see any log even when intentionally trying to ftp some files into this folder. It seems that my previous explanation of the problem was not so clear. Sorry again . What can I do to find an appropriate trace method for my case in your opinion ? _______________________________________________ CentOS mailing list CentOS@xxxxxxxxxx http://lists.centos.org/mailman/listinfo/centos