Sergio Gutierrez wrote: > Hi Tom > > Thank you very much for your answer. > > I'd advise against commenting out the line. What happens if someone > installs the > executable at a later date? > > A common solution is to test to see if the command exists and only run > it if it does. > > [th199096 at warlock ~]> sudo crontab -l | grep "\-x" > 30 3 * * * [ -x /usr/lib/gss/gsscred_clean ] && > /usr/lib/gss/gsscred_clean > 1 2 * * * [ -x /usr/sbin/rtc ] && /usr/sbin/rtc -c > /dev/null 2>&1 > > > So, we could put in code the same thing as the suggested workaround. > But, could we consider it valid for any executable programmed in the > crontab? >
In theory, yes, we could do it for any executable programmed into the crontab. In general, an entry should only be put in the crontab if and only if the executable is installed. Another approach here is to make the addition of the crontab entry be part of the package which contains nfsfind. Right now that is occuring as part of some other process. This solution is cleaner in that it puts the fix where it needs to go. The "-x" is more robust, in case the script gets deleted. > In case it be valid, do we need to implement a kind of notification > for this situation? > No, I don't think we need to do so. > > Thanks, > Tom > > > PS: I'll be willing to sponsor you once you have a fix ready.... > > > > Thanks a lot!. I will be working on it, and I will be posting you. > > Regards. > > Sergio. > >
