GitLab

Implement an exception catching clause, which will respond to any
remaining exception not explicitly handled. In the past there occurred
cases, where aborts went unnoticed. This default clause works well in
csbackchk.py

The change is in response to ticket:244 at BFO.

systemd requires
/usr/bin/systemctl reload syslog.service > /dev/null

This script is specific to BFO and therefor is moved to
code/dataarchive/csback/production in the BFO repository

Merge branch 'master' of ssh://gitserver//home/tforb/git.git/csback

Conflicts:
	BFO2LSDF.sh

The shell script tried to close the VPN connection after having accomplished
copying data files. It still used the Juniper VPN client, which is no longer
in use. I see no reason for closing the VPN connection explicitely. For this
reason the script now leaves the tunnel open.

The juniper VPN client is no longer in use.

BFO2LSDF uses /bin/ip to check for an active VPN tunnel. It reports the IP
address to the system logger. /bin/ip reports the IPv4 and the IPv6 address,
which produced a line break in the output of the previous version of BFO2LSDF,
which in turn confused the log-file parser in csbackntfy. The current version
only outputs the IPv4 address.

Let openvpn establish the connection before testing for an existing
connection.

- update syslog-ng configuration file to be applicable to syslog-ng 3.7.1
- place some notes regarding syslog service in README

A recent syslog-ng (OpenSuSE Tumpleweed, syslog-ng 3.7.1) complained because
of the strange and conflicting source definitions used here. In my opinion all
messages are simply passed internally. No need to tcp or /dev/log. Anyway, the
quickfix here makes things go well.

binary installation directory is hard-wired;
use variable name CSBACKBINDIR instead of LOCBINDIR to avoid conflicts with
Seitosh

increment operator was not working as expected

They way a list of dictionary keys were extracted apparently was not correct