Hello @Mantas,
Sorry for the delay. A full reboot is an option, just requires a bit more preparation.
I made sure plugins.security.restapi.roles_enabled: [all_access, security_rest_api_access]
is present on all nodes, disabled shard allocation and restarted OpenSearch on all nodes.
After all nodes rejoined I reenabled shard allocation, waited for cluster to reach green. Reran securityadmin.sh, restarted Dashboards, alas, still no Security
Afterwards I checked logs to confirm a full cluster restart occured. Seems like it did. So I am unsure still what is wrong.
[INFO ][o.o.c.s.MasterService ] [os-master-03] elected-as-cluster-manager ([2] nodes joined)[{os-master-01}
[INFO ][o.o.c.s.MasterService ] [os-master-03] node-join[{os-master-02}
[INFO ][o.o.c.s.ClusterApplierService] [os-master-03] added {{os-master-02}
[INFO ][o.o.c.s.MasterService ] [os-master-03] node-join[{os-data-02}
[INFO ][o.o.c.s.ClusterApplierService] [os-master-03] added {{os-data-02}
[INFO ][o.o.c.s.MasterService ] [os-master-03] node-join[{os-data-01}
[INFO ][o.o.c.s.ClusterApplierService] [os-master-03] added {{os-data-01}
[INFO ][o.o.c.s.MasterService ] [os-master-03] node-join[{os-data-04}
[INFO ][o.o.c.s.ClusterApplierService] [os-master-03] added {{os-data-04}
[INFO ][o.o.c.s.MasterService ] [os-master-03] node-join[{os-data-03}
[INFO ][o.o.c.s.ClusterApplierService] [os-master-03] added {{os-data-03}