Timo_Lee writes

I am running "recollindex -c /path/to/configurefile" on a directory with a large number of files. It takes too long and hasn’t finished yet.

  1. recollindex outputs the name for each file it has processed. Not sure if it will be faster if such output can be eliminated? But I still want any error report to be printed out.

  2. Also if I ctrl+C the current running recollindex process, will it resume where it is interrupted the next time when I run "recollindex -c /path/to/configurefile" again? I don’t want it to start it over again.

Thanks!

medoc writes

You can set the output file name and verbosity level from the GUI configuration dialog (or have a look at the configuration section in the doc, you can also edit the config file). I don’t think that printing the indexed file names can have any impact on performance except if you do this on a 1200 baud modem…

An interrupted process will resume indexing at the point it was interrupted. The file system walk and up-to-date checks will restart from the start though (but this is reasonably fast).

medoc writes

No actual issue here

Timo_Lee writes

Thanks! Will recollindex log the files it failed to index into some log file?

medoc writes

Yes, among the others. Mostly you want to look at the :x: at the start of lines. x = 2 is for error lines, 3 info, 4 debug/trace.