[ANNOUNCE] OpenSSI 1.0.0 released!!

From: Aneesh Kumar K.V (aneesh.kumar_at_hp.com)
Date: 07/31/04

  • Next message: Olivier Galibert: "EHCI / pci power state / suspend annoying interactions" 
    Date:	Sat, 31 Jul 2004 16:51:32 +0530
To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>, opendlm-devel@lists.sourceforge.net, opengfs-users@lists.sourceforge.net, opengfs-devel@lists.sourceforge.net, linux-cluster@redhat.com

    Hi,

    Sorry for the cross post. I came across this on OpenSSI website. I guess
    others may also be interested.

    -aneesh

    The OpenSSI project leverages both HP's NonStop Clusters for Unixware
    technology and other open source technology to provide a full, highly
    available Single System Image environment for Linux.

    Feature list:
    1. Cluster Membership
       * includes libcluster that application can use
    2. Internode Communication

    3. Filesystem
        * support for CFS over ext3, Lustre Lite
        * CFS can be used for the root
        * reopen of files, devices, ipc objects when processes move is supported
        * CFS supports file record locking and shared writable mapped files
    (along with all other standard POSIX capabilities
        * HA-CFS is configurable for the root or other filesystems
    4. Process Management
         * almost all pieces there, including:
               o clusterwide PIDs
               o process migration and distributed rexec(), rfork() and
    migrate() with reopen of files, sockets, pipes, devices, etc.
               o vprocs
               o clusterwide signalling, get/setpriority
               o capabilities
               o distributed process groups, session, controlling terminal
               o surrogate origin functionality
               o no single points of failure (cleanup code to deal with
    nodedowns)
               o Mosix load leveler (with the process migration model from NSC)
               o clusterwide ptrace() and strace
               o clusterwide /proc/<pid>, ps, top, etc.

    5. Devices
       * there is a clusterwide device model via the devfs code
       * each node mounts its devfs on /cluster/node#/dev and bind mounts it
    to /dev so all devices are visible and accessible from all nodes, but by
    default you see only local devices
       * a process on any node can open a device on any node
       * devices are reopened when processes move
       * processes retain a context, even if they move; the context
    determines which node's devices to access by defaul
    6. IPC
       * all IPC objects/mechanisms are clusterwide:
              o pipes
              o fifos
              o signalling
              o message queues
              o semaphore
              o shared memory
              o Unix-domain sockets
              o Internet-domain sockets
      * reopen of IPC objects is there for process movement
      * nodedown handling is there for all IPC objects
    7. Clusterwide TCP/IP
       * HA-LVS is integrated, with extensions
       * extension is that port redirection to servers in the cluster is
    automatic and doesn't have to be managed.
    8. Kernel Data Replication Service
       * it is in there (cluster/ssi/clreg)
    9. Shared Storage
       * we have tested shared FCAL and use it for HA-CFS
    10. DLM
       * is integrated with CLMS and is HA
    11. Sysadmin
       * services architecture has been made clusterwide
    12. Init, Booting and Run Levels
       * system runs with a single init which will failover/restart on
    another node if the node it is on dies
    13. Application Availability
      * application monitoring/restart provided by spawndaemon/keepalive
      * services started by RC on the initnode will automatically restart on
    a failure of the initnode
    14. Timesync
      * NTP for now
    15. Load Leveling
      * adapted the openMosix algorithm
      * for connection load balancing, using HA-LVS
      * load leveling is on by default
      * applications must be registered to load level
    16. Packaging/Install
       * Have source patch, binary RPMs and CVS source options;
       * Debian packages also available via ap-get repository.
       * First node is incremental to a standard Linux install
       * Other nodes install via netboot, PXEboot, DHCP and simple addnode
    command;
    17. Object Interfaces
       * standard interfaces for objects work as expected
       * no new interfaces for object location or movement except for
    processes (rexec(), migrate(), and /proc/pid/goto to move a process)

    -
    To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
    the body of a message to majordomo@vger.kernel.org
    More majordomo info at http://vger.kernel.org/majordomo-info.html
    Please read the FAQ at http://www.tux.org/lkml/

  • Next message: Olivier Galibert: "EHCI / pci power state / suspend annoying interactions"