First order of business, remove pointing the errors to /dev/null and see what it actually says?

I highly recommend NOT redirecting stderr to /dev/null anyway - you want to see when something goes wrong. You of course do not want to see the standard output with the file update, so indeed redirecting that to /dev/null is fine.

Second order of business, make a shellscript that executes commands in order, instead of guessing how many minutes have passed before running discovery -u. Either you run 4 minutes on a possibly wrong database or rrd schema, or you're too fast and run discovery when svn is still running.

Tom

On 12/11/2017 07:46, Edvinas K wrote: