NFS init

NFS init is a method to "install" and run software on embedded frontend computers (FEC).


  1. System boots kernel and initial ramdisk from TFTPServer
  2. ramdisk sets up network and starts system services (for example ssh)
  3. mount TFTPServer:/common/export/nfsinit to /opt/nfsinit
  4. execute all scripts in alphabetical order in /opt/nfsinit/HOSTNAME/


A FEC by default runs a small custom linux setup that provides only basic system functionality. User applications or special hardware drivers are loaded via NFS. After completing the initial ramdisk setup a folder from an NFS server is mounted. By default this is the TFTPServer providing the initial ramdisk and the path /common/export/nfsinit. This can be overwritten using a kernel commandline parameter (nfsinit=server:path) see EmbeddedSystems. The path is mounted read-only. All FEC can read all initscripts from all hosts.

for systems booting from acc5 the default nfsinit path on the aslx2x server is /common/usr/export/nfsinit/HOSTNAME

for systems booting from acc6 the default nfsinit path on the aslx3x server is /common/export/nfsinit/HOSTNAME

If the FEC should run fesa 32bit classes, see also FESA.FECFilesystem and create startscripts for the requested fesa classes in the fesa init folders.

-- ChristophHandel - 07 Oct 2013
Topic revision: r2 - 28 Jan 2015, ChristophHandel
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback