About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog

Documentation / filesystems / nfs / pnfs.txt




Custom Search

Based on kernel version 3.9. Page generated on 2013-05-02 23:06 EST.

1	Reference counting in pnfs:
2	==========================
3	
4	The are several inter-related caches.  We have layouts which can
5	reference multiple devices, each of which can reference multiple data servers.
6	Each data server can be referenced by multiple devices.  Each device
7	can be referenced by multiple layouts.  To keep all of this straight,
8	we need to reference count.
9	
10	
11	struct pnfs_layout_hdr
12	----------------------
13	The on-the-wire command LAYOUTGET corresponds to struct
14	pnfs_layout_segment, usually referred to by the variable name lseg.
15	Each nfs_inode may hold a pointer to a cache of of these layout
16	segments in nfsi->layout, of type struct pnfs_layout_hdr.
17	
18	We reference the header for the inode pointing to it, across each
19	outstanding RPC call that references it (LAYOUTGET, LAYOUTRETURN,
20	LAYOUTCOMMIT), and for each lseg held within.
21	
22	Each header is also (when non-empty) put on a list associated with
23	struct nfs_client (cl_layouts).  Being put on this list does not bump
24	the reference count, as the layout is kept around by the lseg that
25	keeps it in the list.
26	
27	deviceid_cache
28	--------------
29	lsegs reference device ids, which are resolved per nfs_client and
30	layout driver type.  The device ids are held in a RCU cache (struct
31	nfs4_deviceid_cache).  The cache itself is referenced across each
32	mount.  The entries (struct nfs4_deviceid) themselves are held across
33	the lifetime of each lseg referencing them.
34	
35	RCU is used because the deviceid is basically a write once, read many
36	data structure.  The hlist size of 32 buckets needs better
37	justification, but seems reasonable given that we can have multiple
38	deviceid's per filesystem, and multiple filesystems per nfs_client.
39	
40	The hash code is copied from the nfsd code base.  A discussion of
41	hashing and variations of this algorithm can be found at:
42	http://groups.google.com/group/comp.lang.c/browse_thread/thread/9522965e2b8d3809
43	
44	data server cache
45	-----------------
46	file driver devices refer to data servers, which are kept in a module
47	level cache.  Its reference is held over the lifetime of the deviceid
48	pointing to it.
49	
50	lseg
51	----
52	lseg maintains an extra reference corresponding to the NFS_LSEG_VALID
53	bit which holds it in the pnfs_layout_hdr's list.  When the final lseg
54	is removed from the pnfs_layout_hdr's list, the NFS_LAYOUT_DESTROYED
55	bit is set, preventing any new lsegs from being added.
56	
57	layout drivers
58	--------------
59	
60	PNFS utilizes what is called layout drivers. The STD defines 3 basic
61	layout types: "files" "objects" and "blocks". For each of these types
62	there is a layout-driver with a common function-vectors table which
63	are called by the nfs-client pnfs-core to implement the different layout
64	types.
65	
66	Files-layout-driver code is in: fs/nfs/nfs4filelayout.c && nfs4filelayoutdev.c
67	Objects-layout-deriver code is in: fs/nfs/objlayout/.. directory
68	Blocks-layout-deriver code is in: fs/nfs/blocklayout/.. directory
69	
70	objects-layout setup
71	--------------------
72	
73	As part of the full STD implementation the objlayoutdriver.ko needs, at times,
74	to automatically login to yet undiscovered iscsi/osd devices. For this the
75	driver makes up-calles to a user-mode script called *osd_login*
76	
77	The path_name of the script to use is by default:
78		/sbin/osd_login.
79	This name can be overridden by the Kernel module parameter:
80		objlayoutdriver.osd_login_prog
81	
82	If Kernel does not find the osd_login_prog path it will zero it out
83	and will not attempt farther logins. An admin can then write new value
84	to the objlayoutdriver.osd_login_prog Kernel parameter to re-enable it.
85	
86	The /sbin/osd_login is part of the nfs-utils package, and should usually
87	be installed on distributions that support this Kernel version.
88	
89	The API to the login script is as follows:
90		Usage: $0 -u <URI> -o <OSDNAME> -s <SYSTEMID>
91		Options:
92			-u		target uri e.g. iscsi://<ip>:<port>
93					(allways exists)
94					(More protocols can be defined in the future.
95					 The client does not interpret this string it is
96					 passed unchanged as received from the Server)
97			-o		osdname of the requested target OSD
98					(Might be empty)
99					(A string which denotes the OSD name, there is a
100					 limit of 64 chars on this string)
101			-s 		systemid of the requested target OSD
102					(Might be empty)
103					(This string, if not empty is always an hex
104					 representation of the 20 bytes osd_system_id)
105	
106	blocks-layout setup
107	-------------------
108	
109	TODO: Document the setup needs of the blocks layout driver
Hide Line Numbers
About Kernel Documentation Linux Kernel Contact Linux Resources Linux Blog
Information is copyright its respective author. All material is available from the Linux Kernel Source distributed under a GPL License. This page is provided as a free service by mjmwired.net.