Package issues during update

It was found that when updating to Stork 1.15.0 (in this case from 1.14.0) on RHEL 7 (and CentOS 7 though engineering states that this problem likely applies to all versions) that there are two problems encountered:

1. The Stork Server service ends up stopped in the disabled state even though it was enabled and started prior to update.
2. The useradd call in the postinstall script sets the homedir to /var/lib instead of /var/lib/stork-server.

A patch was provided by the reporter:

diff --git a/etc/hooks/rpm/isc-stork-server.postinst b/etc/hooks/rpm/isc-stork-server.postinst
index 3b890b75..7833efd4 100644
--- a/etc/hooks/rpm/isc-stork-server.postinst
+++ b/etc/hooks/rpm/isc-stork-server.postinst
@@ -4,5 +4,5 @@ set -eu

# add stork-server user if does not exist
if ! getent passwd stork-server > /dev/null; then
- useradd --system --home-dir /var/lib/ stork-server
+ useradd --system --base-dir /var/lib/ stork-server
fi
diff --git a/etc/hooks/rpm/isc-stork-server.prerm b/etc/hooks/rpm/isc-stork-server.prerm
index e4649e2c..cc007fbc 100644
--- a/etc/hooks/rpm/isc-stork-server.prerm
+++ b/etc/hooks/rpm/isc-stork-server.prerm
@@ -1,16 +1,17 @@
#!/bin/sh

set -eu
-
-has_active_systemd=0
-if command -v systemctl > /dev/null; then
- status=$(systemctl is-system-running || true)
- if [ "${status}" = "running" ] || [ "${status}" = "degraded" ] || [ "${status}" = "maintenance" ]; then
- has_active_systemd=1
+if [ "$1" -eq 0 ]; then # Uninstall == 0 not Upgrade == 1
+ has_active_systemd=0
+ if command -v systemctl > /dev/null; then
+ status=$(systemctl is-system-running || true)
+ if [ "${status}" = "running" ] || [ "${status}" = "degraded" ] || [ "${status}" = "maintenance" ]; then
+ has_active_systemd=1
+ fi
fi
-fi

-if [ $has_active_systemd -eq 1 ]; then
- systemctl disable isc-stork-server
- systemctl stop isc-stork-server
-fi
+ if [ $has_active_systemd -eq 1 ]; then
+ systemctl disable isc-stork-server
+ systemctl stop isc-stork-server
+ fi
+fi

SF1727