Samisa Abeysinghe <[EMAIL PROTECTED]> writes: >> AFAIK, that file is backed up (axis2.log.1) and start new file >> (axis2.log). Then , if new one also exceeds 1M, it is backed up over >> writing old backed up file. >> >> Finally, system has only two files, current log file (axis2.log) and >> immediate >> recent log file (axis2.log.1). >> > So it is only two levels. Is that good enough? Good enough than a single file , I guess :)
> The other concern is that, the code is already commited and it was > never discussed on this list, despite it being a major decision :) +1, I also didn't see any discussion regarding this change beforehand. > Have we implemented this so that it works for all the deployment > models like simple http/tcp servers as well as the httpd module? Hope, Damitha will chime in some point :) >> >>> Also, for a production system, I deem that 1M is too small. >>> >> I guess not :), I looked at some other logs (#ls -lah >> /var/log/ ) they even hardly exceeds 300K. >> > I guess you looked on your PC, it is not a production system, it is? ;-) > On my PC, apache error log is 491K already and I hardly run it on a > daily basis. I'm not sure whether apache using rotational or not. My point is since log file is a text file, even 1M is too large. WDT? thanks, Dinesh --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
