We ran into an issue when patching our RHEL9 servers. As they were patched, the /etc/init.d directory was removed. And as it turns out, XL Deploy still relies on this directory to exist to be able to run.
System V has been depecrated for a while, starting in RHEL7, which itself is out of support. System V support is now actively being hindered when patching, as the /etc/init.d directory was removed. The application is still able to run as long as the directory is recreated, but moving to a systemd supported installation would and should be preferred.
Moving to RHEL10 it looks like support for system V will be completely removed, so looking at future proofing the application this should be looked at.
links:
- https://linuxjust4u.com/linux/systemd-260-sysv-removed-linux-upgrade-guide/
- https://docs.redhat.com/en/documentation/red_hat_enterprise_linux/10/pdf/10.0_release_notes/Red_Hat_Enterprise_Linux-10-10.0_Release_Notes-en-US.pdf
by: Simon v. | about a month ago | Installation & Upgrades
Comments
Thank you for raising this.
We have identified the /etc/init.d dependency in the legacy host-service installation path and plan to replace it with native systemd support in a future release.
The dependency is specific to the host-based YAJSW service installation flow invoked through install-service.sh. Containerized deployments should normally start Deploy directly as the container process and do not use this service installation path.
To confirm that this is the installation path you are using, can you let us know whether Deploy is installed as a native OS daemon on the host — using an extracted tarball with install-service.sh to register it as a service — or whether it is running as a Docker /container image?