Changes in Swarm Cluster Services (SCS) 1.6

This release features the following improvements:

• Support for static IPs is further enhanced, with the ability to set specific addresses manually. (SCSL-363)

• Administrative user names other than “admin” are supported. Additionally, if storage nodes are unavailable to receive the administrative update, the update will fail. However, in circumstances where the update should be permitted, a new flag (--allow-notification-failures) has been added to the scsctl {component} config set .... See the help text for additional details. (SCSL-351 and SCSL-355)

• When an instance must be removed from SCS, metadata and in-memory caches are completely cleared of the reference. (SCSL-362)

• Every time DHCP is initialized, a backup copy of the prior dhcpd.conf file will be created, and its location will be printed on the screen. This will help with migrating existing manual updates. (SCSL-357)

• In cases where a sysctl file needs to be provided to storage nodes, a new static file name of SYSCTL is supported for this purpose. Swarm Storage 15.3.1 is also required for this. (SCSL-264)

• Improvements around systemd have been included. However, SELinux must be disabled when initializing or upgrading SCS. (SCSL-242, SCSL-365, and SCSL-367)

• The prompt for the initial administrative password has been moved into the scsctl init wizard command. (SCSL-360)

• Swarm Storage authentication is no longer broken when the administrative password is changed from its existing value to the same value. (SCSL-364).

• Detailed logs from the initialization wizard or upgrade process have been moved to the main logging folder (/var/log/datacore/) (SCSL-371).

Limitations

• SELinux must be disabled when initializing or upgrading SCS. (SCSL-365)

• Performing a system reset (scsctl system reset) may result in an error if it runs after an upgrade. Contact support if this occurs. (SCSL-370)

• There is no support for using gPXE instead of iPXE in this release. (PLT-60)