Many process blocked fore more than 120 seconds, even ps command

Hi!
Since upgrading to RHEL 6.3 (Linux version 2.6.32-279.el6.x86_64) I regularly get some "blocked for more than 120 seconds" messages, with some processes locked and unresponsive.
It happens randomly on process that are writing on NFS devices. I thought it may come from NFS deviece problem but recently I saw that even the "ps" command can be blocked (and indeed never returns)!

Is this a known issue ? Is there a fix ?

Many thanks,
Xavier.

Here is an example:
Feb 19 21:28:33 svr428 kernel: ps D 0000000000000000 0 8914 8896 0x00000080
Feb 19 21:28:33 svr428 kernel: ffff880197c85c50 0000000000000086 ffff8801861480c0 ffffffff8161f700
Feb 19 21:28:33 svr428 kernel: ffff88017f7c0ab8 ffff880197c85bd8 ffffffff811e3a83 ffff8801861480c0
Feb 19 21:28:33 svr428 kernel: ffff8801982c65f8 ffff880197c85fd8 000000000000fb88 ffff8801982c65f8
Feb 19 21:28:33 svr428 kernel: Call Trace:
Feb 19 21:28:33 svr428 kernel: [] ? proc_pident_instantiate+0x93/0xc0
Feb 19 21:28:33 svr428 kernel: [] rwsem_down_failed_common+0x95/0x1d0
Feb 19 21:28:33 svr428 kernel: [] rwsem_down_read_failed+0x26/0x30
Feb 19 21:28:33 svr428 kernel: [] call_rwsem_down_read_failed+0x14/0x30
Feb 19 21:28:33 svr428 kernel: [] ? down_read+0x24/0x30
Feb 19 21:28:33 svr428 kernel: [] __access_remote_vm+0x41/0x1f0
Feb 19 21:28:33 svr428 kernel: [] ? security_inode_permission+0x1f/0x30
Feb 19 21:28:33 svr428 kernel: [] ? nameidata_to_filp+0x54/0x70
Feb 19 21:28:33 svr428 kernel: [] access_process_vm+0x5b/0x80
Feb 19 21:28:33 svr428 kernel: [] proc_pid_cmdline+0x6d/0x120
Feb 19 21:28:33 svr428 kernel: [] ? alloc_pages_current+0xaa/0x110
Feb 19 21:28:33 svr428 kernel: [] proc_info_read+0xad/0xf0
Feb 19 21:28:33 svr428 kernel: [] vfs_read+0xb5/0x1a0
Feb 19 21:28:33 svr428 kernel: [] ? audit_syscall_entry+0x272/0x2a0
Feb 19 21:28:33 svr428 kernel: [] sys_read+0x51/0x90
Feb 19 21:28:33 svr428 kernel: [] system_call_fastpath+0x16/0x1b