File Systems FAQ

• Can I run one or more of the cluster nodes as highly available NFS server(s) with other cluster nodes as clients?

  No. Issues exist with local locking interferes having the ability to kill and restart lockd (which occurs during NFS failover). Between the kill and restart, a blocked local process can be granted the lock, which prevents the client system that owns the lock from reclaiming it after failover.

• Can I use a cluster file system for applications that are not under Resource Group Manager control?

  Yes. However, without RGM control, the applications cannot survive the failure of the node on which they are running.

• Must all cluster file systems have a mount point under the /global/device-group directory?

  No. However, placing cluster file systems under the same mount point, such as /global/device-group, enables better organization and management of these file systems.

• What are the differences between using the cluster file system and exporting NFS file systems?

  There are several differences:

  1. The cluster file system supports global devices. NFS does not support remote access to devices.

  2. The cluster file system has a global namespace. Only one mount command is required. With NFS, you must mount the file system on each node.

  3. The cluster file system caches files in more cases than does NFS. For example when a file is being accessed from multiple nodes for read, write, file locks, async I/O.

  4. The cluster file system supports seamless failover if one server fails. NFS supports multiple servers, but failover is only possible for read-only file systems.

  5. The cluster file system is built to exploit future fast cluster interconnects that provide remote DMA and zero-copy functions.

  6. If you change the attributes on a file (using chmod(1M), for example) in a cluster file system, the change is reflected immediately on all nodes. With an exported NFS file system, this can take much longer.