US007707263B1
`
`(12) Ulllted States Patent
`Cramer et a].
`
`(10) Patent N0.:
`(45) Date of Patent:
`
`US 7,707,263 B1
`Apr. 27, 2010
`
`(54) SYSTEM AND METHOD FOR ASSOCIATING
`A NETWORK ADDRESS WITHA STORAGE
`
`6,591,306 B1 *
`6,621,820 B1 *
`
`7/2003 Redlich .................... .. 709/245
`9/2003 Williams 6161. ..... .. 370/39531
`
`DEVICE
`
`6,636,499 B1 * 10/2003 DoWling ................... .. 370/338
`
`(75) Inventors: Samuel M- Cramer’ Sunnyvale’ CA
`
`6,718,383 B1 *
`
`4/2004 Hebert ..................... .. 709/224
`
`
`
`Joydeep sen sarma’ MountainvieW, CA (US); Richard 0.
`
`
`
`* 6,810,396 B1 * 10/2004 Blumenau et a1. ............ .. 707/5 Blumenau et a1. . . . . . . . . . ..
`
`
`
`Larson, CuPemnO, CA (US)
`
`6,870,852 B1 *
`
`3/2005 Lawitzke .................. .. 370/401
`
`(73) Assignee: NetApp, Inc., Sunnyvale, CA (US)
`
`7/2005 Cmmer et 31'
`6’920’580 Bl
`7,079,499 B1* 7/2006 Akhtar et a1. ............. .. 370/310
`
`( * ) Notice:
`
`Subject to any disclaimer, the term of this
`patent is extended or adjusted under 35
`USC 154(b) by 764 days.
`
`(21) Appl. No.: 10/138,918
`
`(22) Filed:
`(51) Int Cl
`
`May 3’ 2002
`
`(2006.01)
`G06F 15/16
`(52) US. Cl. ..................... .. 709/212; 709/203; 709/201;
`709/223; 709/220; 709/204
`(58) Field of Classi?cation Search ............... .. 709/203,
`709/201’ 21’ 220’ 223’ 204’ 212
`See application ?le for complete search history.
`
`(56)
`
`References Cited
`
`U.S. PATENT DOCUMENTS
`
`5,163,131 A 11/1992 ROW et a1~
`5’355’453 A 10/1994 R(_)W et 31'
`5,485,579 A
`1/1996 HltZ et a1.
`5 802 366 A
`9/l998 ROW et 31‘
`538193292 A l 0 / 1998 HitZ et al‘
`5,854,901 A * 12/1998 C016 et a1. ................ .. 709/245
`5,931,918 A
`8/1999 ROW et a1‘
`5,941,972 A
`8/1999 Hoese et a1,
`5,963,962 A 10/ 1999 Hitz et a1.
`6,065,037 A
`5/2000 HitZ et a1~
`6,243,759 B1 *
`6/2001 Boden et a1. .............. .. 709/238
`6,289,356 B1
`9/2001 HitZ et a1.
`6,425,035 B2
`7/2002 Hoese et a1.
`6,574,667 B1 *
`6/2003 Blumenau et a1. ......... .. 709/229
`6,581,102 B1 *
`6/2003 Amini et a1. .............. .. 709/231
`
`_
`(Cont1nued)
`OTHER PUBLICATIONS
`
`Request for Comments (RFC) 826: An Ethernet Address Resolution
`Protocol, Internet Engineering Task Force (IFTF), Nov. 1982.
`
`(Continued)
`
`_
`_
`Prlmary ExammeriThu Ngqyen
`ASS/SlamExammeriLan-Dal T Truong
`(74) Attorney, Agent, or Firm4Cesari and McKenna, LLP
`
`(57)
`
`ABSTRACT
`
`A system and method for associating a network address With
`a volume, individual disk or collection of disks in a netWork
`storage system. Identifying information is stored on each
`.
`volume or d1sk so that a ?le server can map a netWork address
`to a MAC address associated With a particular netWork inter
`face Controller Of a ?le server- Input/01111311t Operations
`directed to the netWork address associated With a particular
`volume or disk is directed to the ?le server that is currently
`managing that volume or disk. In a system utilizing these
`address associated volumes, the name and address of data
`does not Change as different ?le Servers manage a particular
`Volume or disk'
`
`13 Claims, 11 Drawing Sheets
`
`918541181415“ i 1
`
`400
`
`S
`404
`|
`THISAN lg VOLUME
`
`40 m I
`
`OBTAIN IP ADDRESS
`FOR STORAGE DEVICE
`I
`410" MAP IP ADDRESS TO NIC
`I
`412‘ ADVERTISE lP ADDRESS
`OVER APPROPRIATE NICE
`
`416
`J
`
`COMPLETE
`
`"0
`
`414
`MORE
`DEVICESTO
`
`CONFI?GURE
`
`YES
`
`LENOVO ET AL. EXHIBIT 1005
`Page 1 of 20
`
`

`
`US 7,707,263 B1
`Page 2
`
`U.S. PATENT DOCUMENTS
`
`2003/0126118 A1*
`2003/0131207 A1*
`2003/0177206 A1*
`2006/0265529 A1*
`2007/0012931 A1*
`2007/0112931 A1*
`
`7/2003
`7/2003
`9/2003
`11/2006
`1/2007
`5/2007
`
`Burton et al. ................ .. 707/3
`Arakawa et al.
`711/162
`Whitlow ................... .. 709/220
`
`Kuik et a1. .................. .. 710/62
`
`Kuik et a1. ................ .. 709/216
`
`OTHER PUBLICATIONS
`
`Common Internet File System (CIFS) Version: CIFS-Spec 0.9, Stor
`age Networking Industry Association (SNIA), Draft SNIA CIFS
`Documentation Work Group Work-In-Progress, Revision Date: Mar.
`26, 2001‘
`
`7,296,068
`2002/0023150
`2002/0120706
`2002/0138628
`2002/0143946
`2002/0147774
`2002/0165906
`2003/0023784
`2003/0088700
`2003/0101109
`2003/0115324
`2003/0120743
`
`B1
`A1 *
`
`11/2007 Sarma et a1.
`2/2002 Osafune et a1. ...... ..
`
`8/2002 Murphy ....... ..
`A1 *
`9/2002 Tingley et a1. .
`A1 *
`A1 * 10/2002 Crosson .... ..
`A1 * 10/2002 Lisiecki et al. .
`
`A1* 11/2002 Ricart et a1. ......... ..
`
`709/221
`709/208
`. 709/227
`
`. 709/226
`709/203
`709/203
`
`
`
`A1 * A1 *
`
`
`
`1/2003 Matsunami et al. 5/2003 Aiken ............. ..
`
`
`
`A1 * A1 *
`
`
`
`5/2003 Kaneda et al. 6/2003 Blumenau et al. .... ..
`
`..... .. 710/36
`
`. 709/245
`705/28
`709/225
`
`A1
`
`6/2003 Coatney et al.
`
`* cited by examiner
`
`LENOVO ET AL. EXHIBIT 1005
`Page 2 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 1 0f 11
`
`US 7,707,263 B1
`
`109
`
`INTERNET
`
`107,
`NETWORK
`CACHE
`
`104,
`
`V
`PCS J
`l
`i -
`
`100\
`
`E
`SERVERS
`
`108\
`’
`SWITCH,
`ROUTER
`
`LAN
`m
`
`100
`/
`
`‘
`
`MGMT
`CONSOLE
`@ ,
`
`F|LER1
`
`11°
`
`F|LER2
`
`112
`
`,
`
`120
`
`SWITCHING
`NETWORK
`
`122
`
`122
`
`122
`
`122
`
`FIG. 1
`
`LENOVO ET AL. EXHIBIT 1005
`Page 3 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 2 0f 11
`
`US 7,707,263 B1
`
`MEQQZRY
`——
`
`PROCESSOR
`22
`
`NVRAM
`262
`
`110
`J
`
`/ 225
`
`NETWORK
`ADAPTER
`2.2.2
`
`,
`L__;_
`TO/FROM
`LAN 102
`
`STORAGE
`ADAPTER
`2.8.
`
`i
`
`TO/FROM
`SWITCHING tSJETWORK
`12
`
`FIG. 2
`
`LENOVO ET AL. EXHIBIT 1005
`Page 4 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 3 0f 11
`
`US 7,707,263 B1
`
`8~\
`
`./
`
`[Ill 0
`
`@z_n_n_<_>_ _ T P
`
`|...I| _ III.
`
`_ A _
`/ Q \
`
`- - lull
`
`/ \
`
`2%? _ T: N;
`
`QTLwEmE _ M _ mm mm
`03 _ H " $6 ,5:
`M2105 _ n 0 a8
`
`[Ill Ill
`
`/ \ 1 ,
`
`155 _
`$210 " _ $52 $52
`
`% u _ _ u am
`
`_
`
`5,5 555 BE
`ow.
`
`m .wE
`
`
`
`Q55 5E6: 210252 5E6:
`
`
`
`
`
`LENOVO ET AL. EXHIBIT 1005
`Page 5 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 4 0f 11
`
`US 7,707,263 B1
`
`402
`
`IDENTIFY NEXT DISK
`TO CONFIGURE
`
`404
`IS
`THIS AN IE; VOLUME
`
`YES
`
`408w
`
`OBTAIN IP ADDRESS
`FOR STORAGE DEVICE
`
`410*“ MAP IP ADDRESS TO NIC
`
`412‘ ADVERTISE IP ADDRESS
`OVER APPROPRIATE NICS
`
`MORE
`DEVICES TO
`CONFIIPGURE
`
`YES
`
`FIG. 4
`
`LENOVO ET AL. EXHIBIT 1005
`Page 6 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 5 0111
`
`US 7,707,263 B1
`
`(500
`
`,ENG
`% ,MKT
`
`E0’
`
`E1 ’
`
`E2’
`
`E4A7
`
`E413’
`
`502‘
`
`FlLER1
`110/
`
`FILER 2
`112’
`
`5047
`
`voua
`
`510
`
`515
`
`FIG. 5
`
`LENOVO ET AL. EXHIBIT 1005
`Page 7 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 6 0f 11
`
`US 7,707,263 B1
`
`502
`
`504
`
`602* ENG-NiCS
`
`MKT-NICS
`
`FIG. 6
`
`7o2- ENG-NICS
`
`704* MKT-NICS
`
`E4B
`
`E4A
`
`FIG. 7
`
`LENOVO ET AL. EXHIBIT 1005
`Page 8 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 7 0f 11
`
`US 7,707,263 B1
`
`VDLUME
`NAME
`ggg
`
`NETWORK
`ADDRESS
`s_1g
`
`515
`‘/
`
`LOGICAL
`INTERFACES
`TO BE
`ADVERTISED 0N
`_a_1_5
`
`FIG. 8
`
`LENOVO ET AL. EXHIBIT 1005
`Page 9 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 8 0f 11
`
`US 7,707,263 B1
`
`f 900
`
`CLIENTS
`em
`
`ENG-NICS 3
`
`Q
`
`I’ 910
`
`E_1
`
`MKT-NICS
`i
`
`E2
`
`SWITCHING
`
`FIG. 9
`
`LENOVO ET AL. EXHIBIT 1005
`Page 10 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 9 0f 11
`
`US 7,707,263 B1
`
`1005
`
`ENG-NICS
`
`E0
`
`1000
`
`FIG. 10
`
`LENOVO ET AL. EXHIBIT 1005
`Page 11 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 10 0f 11
`
`US 7,707,263 B1
`
`1105 1 ENG-NICS
`
`1110 4 MKT-NICS
`
`E1
`
`E2
`
`FIG. 11
`
`LENOVO ET AL. EXHIBIT 1005
`Page 12 of 20
`
`

`
`US. Patent
`
`Apr. 27, 2010
`
`Sheet 11 0111
`
`US 7,707,263 B1
`
`1205*
`
`VOL1
`
`1. 2. 3. 4
`
`ENG-NICS
`
`FIG. 12
`
`LENOVO ET AL. EXHIBIT 1005
`Page 13 of 20
`
`

`
`US 7,707,263 B1
`
`1
`SYSTEM AND METHOD FOR ASSOCIATING
`A NETWORK ADDRESS WITH A STORAGE
`DEVICE
`
`FIELD OF THE INVENTION
`
`The present invention relates to networked storage sys
`tems, and more particularly to accessing ?les in a networked
`storage system.
`
`BACKGROUND OF THE INVENTION
`
`A ?le server is a computer that provides ?le service relating
`to the organiZation of information on storage devices, such as
`disks. The ?le server or ?ler includes a storage operating
`system that implements a ?le system to logically organiZe the
`information as a to hierarchical structure of directories and
`?les on the disks. Each “on-disk” ?le may be implemented as
`a set of data structures, e.g., disk blocks, con?gured to store
`information. A directory, on the other hand, may be imple
`mented as a specially formatted ?le in Which information
`about other ?les and directories are stored.
`A ?ler may be further con?gured to operate according to a
`client/ server model of information delivery to thereby alloW
`many clients to access ?les stored on a server, e. g., the ?ler. In
`this model, the client may comprise an application, such as a
`database application, executing on a computer that “con
`nects” to the ?ler over a computer netWork, such as a point
`to-point link, shared local area netWork (LAN), Wide area
`netWork (WAN), or virtual private netWork (V PN) imple
`mented over a public netWork such as the Internet. Each client
`may request the services of the ?le system on the ?ler by
`issuing ?le system protocol messages (in the form of packets)
`to the ?ler over the netWork.
`A common type of ?le system is a “Write in-place” ?le
`system, an example of Which is the conventional Berkeley fast
`?le system. In a Write in-place ?le system, the locations of the
`data structures, such as Modes and data blocks, on disk are
`typically ?xed. An Mode is a data structure used to store
`information, such as meta-data, about a ?le, Whereas the data
`blocks are structures used to store the actual data for the ?le.
`The information contained in an inode may include, e.g.,
`oWnership of the ?le, access permission for the ?le, siZe of the
`?le, ?le type and references to locations on disk of the data
`blocks for the ?le. The references to the locations of the ?le
`data are provided by pointers, Which may further reference
`indirect blocks that, in turn, reference the data blocks,
`depending upon the quantity of data in the ?le. Changes to the
`inodes and data blocks are made “in-place” in accordance
`With the Write in-place ?le system. If an update to a ?le
`extends the quantity of data for the ?le, an additional data
`block is allocated and the appropriate inode is updated to
`reference that data block.
`Another type of ?le system is a Write-anyWhere ?le system
`that does not overWrite data on disks. If a data block on disk
`is retrieved (read) from disk into memory and “dirtied” With
`neW data, the data is stored (Written) to a neW location on disk
`to thereby optimiZe Write performance. A Write-anyWhere ?le
`system may initially assume an optimal layout such that the
`data is substantially contiguously arranged on disks. The
`optiis mal disk layout results in ef?cient access operations,
`particularly for sequential read operations, directed to the
`disks. A particular example of a Write-anyWhere ?le system
`that is con?gured to operate on a ?ler is the Write AnyWhere
`File Layout (WAFLTM) ?le system available from NetWork
`Appliance, Inc. of Sunnyvale, Calif. The WAFL ?le system is
`implemented Within a microkernel as part of the overall pro
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`55
`
`60
`
`65
`
`2
`tocol stack of the ?ler and associated disk storage. This micro
`kernel is supplied as part of NetWork Appliance’s Data
`ONTAPTM softWare, residing on the ?ler, that processes ?le
`service requests from netWork-attached clients.
`As used herein, the term “storage operating system” gen
`erally refers to the computer-executable code operable on a
`storage system that manages data access, and, in the case of
`?lers, implements ?le system semantics (such as the above
`referenced WAFL). In this sense, ONTAP softWare is an
`example of such a storage operating system implemented as a
`microkernel. The storage operating system can also be imple
`mented as an application program operating over a general
`purpose operating system, such as UNIX or WindoWs NT®,
`or as a general-purpose operating system With con?gurable
`functionality, Which is con?gured for storage applications as
`described herein.
`Disk storage is typically implemented as one or more stor
`age “volumes” that comprise physical storage disks, de?ning
`an overall logical arrangement of storage space. Currently
`available ?ler implementations can serve a large number of
`discrete volumes (150 or more, for example). Each volume is
`associated With its oWn ?le system and, for purposes hereof,
`volume and ?le system shall generally be used synony
`mously. The disks Within a volume are typically organiZed as
`one or more groups of Redundant Array of Independent (or
`Inexpensive) Disks (RAID). RAID implementations enhance
`the reliability/ integrity of data storage through the redundant
`Writing of data “stripes” across a given number of physical
`disks in the RAID group, and the appropriate caching of
`parity information With respect to the striped data. In the
`example of a WAFL ?le system, a RAID 4 implementation is
`advantageously employed. This implementation speci?cally
`entails the striping of data across a group of disks, and sepa
`rate parity caching Within a selected disk of the RAID group.
`As described herein, a volume typically comprises at least
`one data disk and one associated parity disk (or possibly
`data/parity) partitions in a is single disk) arranged according
`to a RAID 4, or equivalent high-reliability, implementation.
`More than one ?ler can reside on a single netWork (LAN,
`WAN, etc.), for access by netWork-connected clients and
`servers. Where multiple ?lers are present on the netWork,
`each ?ler may be assigned responsibility for a certain set of
`volumes. The ?lers may be connected in a cluster using a
`separate physical interconnect or linking communication pro
`tocol that passes over the netWork (eg the LAN, etc.). In the
`event of a failure or shutdoWn of a given ?ler, its volume set
`can be reassigned to another ?ler in the cluster to maintain
`continuity of service. In the case of a shutdoWn, various
`failover techniques are employed to preserve and restore ?le
`service, as described generally in commonly oWned US.
`patent application Ser. No. 09/933,883 entitled OPERATED
`INITIATED GRACEFUL TAKEOVER IN A NODE CLUS
`TER by Naveen Bali et al and US. patent application Ser. No.
`09/625,234 entitled NEGOTIATING TAKEOVER IN HIGH
`AVAILABILITY CLUSTER by Samuel M. Cramer et al., the
`teachings of Which are expressly incorporated herein by ref
`erence. Such techniques involve (a) the planned and
`unplanned takeover of a ?ler’s volumes by a cluster partner
`?ler upon ?ler shutdoWn; and (b) the giveback of the taken
`over volumes by relinquishing control by the cluster partner
`?ler. A management station can also reside on the netWork, as
`a specialiZed client that includes storage management soft
`Ware used by a system administrator to manipulate and con
`trol the storage-handling by netWorked ?lers.
`A ?ler can be made more reliable and stable in the event of
`a system shutdoWn or other unforeseen problem by employ
`ing a backup memory consisting of a non-volatile random
`
`LENOVO ET AL. EXHIBIT 1005
`Page 14 of 20
`
`

`
`US 7,707,263 B1
`
`3
`access memory (NV RAM). An NVRAM is typically a large
`volume solid-state memory array (RAM) having either a
`backup battery, or other built-in last-state-retention capabili
`ties (eg a FLASH memory), that holds the last state of the
`memory in the event of any poWer loss to the storage system.
`A ?ler is typically made more reliable and stable in the
`event of a system shutdown or unforeseen problem by
`employing a backup memory consisting of a non-volatile
`random access memory NVRAM. An NVRAM is typically a
`large-volume, solid-state memory array (RAM) having either
`a backup battery, or other built-in last-state-retention capa
`bilities (e. g. a FLASH memory), that holds the last state of the
`memory in the event is of any poWer loss to the storage
`system. In a knoWn implementation, each client transaction
`request processed by the storage operating system is logged
`to the NVRAM as a journal entry. The NVRAM is loaded
`With requests until such time as a consistency point (CP) is
`reached. CPs occur at ?xed time intervals, or When other key
`events arise. In the event of a fault, poWer loss, or other
`interruption in the normal How of information among the
`20
`client, storage operating system, and the disks, the NVRAM
`log is replayed to re-perform any requests logged therein for
`its oWn ?ler (and an associated cluster partner ?ler, if any)
`betWeen the last CP and an interruption in storage handling.
`In addition, the log is replayed during reboot. Each time a CP
`occurs, the requests logged in the NVRAM are subsequently
`overWritten or otherWise cleared, once the results of the
`requests are Written from the ?ler’ s conventional RAM buffer
`cache to disk. Immediately thereafter, the NVRAM is avail
`able for the logging of neW requests.
`In the event of a shutdoWn, poWer failure or other system
`problem, Which interrupts the normal flow of information
`among the client, storage operating system, and the disks, the
`NVRAM can be used to recover information logged since the
`last CP prior to the interruption event.
`Computer netWorks are typically composed of computers
`that communicate With each other through speci?c protocols.
`Each computer or device attached to a netWork is assigned at
`least one netWork-unique name, i.e. a netWork address. Any
`device in the netWork can manage ?les that are stored on
`devices Which are locally accessible to that device. Addition
`ally, a netWork device could require access to ?les that are
`managed by other netWork devices Within the netWork.
`In knoWn ?le system implementations, local ?lesithose
`?les Which are stored on devices locally accessible to a net
`Work deviceiare identi?ed by a device name and a ?le name.
`For example, in a Microsoft Windows@ compatible device, a
`?le may be identi?ed by “C:/foo/bar/?le.doc.” In this example
`the device is called C:, Which is the name of a disk connected
`to the Microsoft WindoWs machine. The ?le name identi?es
`the directory structure and individual ?le to be accessed. In
`this example, the ?le “?le.doc” is stored in the “bar” subdi
`rectory of the “foo” directory. This combination produces a
`unique ?le identi?er for those ?les Which are locally acces
`sible. Files Which are accessed via the netWork, i.e. non-local
`?les, are identi?ed by the device name and ?le name as local
`?les are, but are further identi?ed by the address of the com
`puter or netWork device that manages that ?le. This combi
`nation of netWork address, device name, and ?le name pro
`duces a netWork-unique identi?er for a particular ?le.
`As part of a netWork storage system, clients request the
`services of a ?ler’s ?le system, i.e. making a data access
`request, by issuing ?le system protocol messages (in the form
`of packets) to the ?ler over the netWork. The ?ler ful?lls a data
`access request by locating and packaging the data, and then
`sending the data through the netWork back to the requesting
`client. In knoWn implementations, the data access request
`
`50
`
`55
`
`60
`
`65
`
`25
`
`30
`
`35
`
`40
`
`45
`
`4
`identi?es the data through an identi?er that contains the net
`Work address of the ?ler, for example “1.2.3.4:C1/foo/bar/
`?le.doc.” In this example, the ?le named “?le.doc” is located
`in the “/foo/bar” directory on drive C: of the ?ler Whose
`netWork address is 1.2.3.4. The data that the client has
`requested could be physically located anyWhere in the net
`Work, on any of the possible storage devices accessible
`through any ?le server.
`A noted disadvantage of prior implementations arises for a
`client When a ?ler is taken out the netWork, and its netWork
`address becomes unavailable to other computers or netWork
`devices connected to the netWork. It is possible to reconnect
`the storage devices that a particular ?le server manages to
`another ?ler, but in doing so, the data identi?cation must
`change to the netWork address of the neW ?ler in order to
`properly locate the data managed by the ?ler Which has been
`taken out of the netWork. This change of the data identi?ca
`tion can be inconvenient and time-consuming, especially
`considering that a ?ler can be taken off of the netWork for
`many reasons, including ?ler upgrade, scheduled mainte
`nance, or unexpected system failure. The ability to transpar
`ently access netWork data before, during, and after storage
`device recon?guration to a neW ?ler could vastly increase
`con?guration ?exibility and data accessibility, and could
`decrease doWntime resulting from ?ler or other server failure.
`In addition, With the groWing demand for electronic ?le
`availability comes the challenge of backing up and restoring
`storage devices that can have extremely large storage capaci
`ties. To reduce the doWntime associated With the back-up/
`restore operation, administrators often resort to imposing
`limits on storage and enforcing quotas. Fasterback-up/restore
`capabilities and more ?exibility in terms of adding storage
`devices can both reduce doWntime and decrease the numbers
`of restrictions imposed on users of the system. Accordingly,
`the potential inaccessibility of data described above is another
`disadvantage that impedes these advantageous goals.
`Accordingly, it is an object of the invention to provide for
`a system and method for associating netWork addresses With
`volumes so that clients can seamlessly and transparently
`address I/O operations to the speci?ed netWork address for a
`given volume Without concern for the netWork address of the
`?le server managing the particular volume.
`
`SUMMARY OF THE INVENTION
`
`This invention overcomes the disadvantages of the prior art
`by providing a system and method for identi?cation of a
`storage device by an IP address, or similar netWorkbased
`address. Thus, each storage device is netWork accessible by
`all devices connected to the netWork Without regard to the
`identity of the storage device’s ?le management system. The
`storage device is associated With one or more IP addresses.
`According to an aspect of the invention, the IP address and
`other con?guration information is stored at a predetermined
`location on the storage device itself for access by ?lers,
`thereby permitting the storage device to be portable among
`various ?lers. Each ?ler or other netWork device in a netWork
`obtains the IP address of an IP volume. I/O operations to that
`volume are then directed to the IP address of the particular
`volume. This IP addressing permits the ?ler that is managing
`a particular volume to be changed so that the transfer is
`transparent to a client accessing data on a volume.
`In accordance With one embodiment of the invention, the
`netWork address of a volume is mapped to the machine
`address code (MAC) address of a netWork interface controller
`(N IC) of the ?le server that is currently managing the volume.
`
`LENOVO ET AL. EXHIBIT 1005
`Page 15 of 20
`
`

`
`US 7,707,263 B1
`
`5
`Thus, I/O requests that are directed to the network address of
`the volume are redirected to the appropriate ?le server for
`processing.
`
`BRIEF DESCRIPTION OF THE DRAWINGS
`
`The above and further advantages of the invention may be
`better understood by referring to the following description in
`conjunction with the accompanying drawings in which like
`reference numerals indicate identical or functionally similar
`elements:
`FIG. 1 is a block diagram of an exemplary network envi
`ronment;
`FIG. 2 is a schematic block diagram of an exemplary ?le
`server;
`FIG. 3 is a schematic block diagram of an exemplary stor
`age operating system;
`FIG. 4 is a ?owchart of the process to con?gure volumes as
`IP volumes in accordance with the invention;
`FIG. 5 is a schematic block diagram of an exemplary net
`work environment in which the principles of the present
`invention are performed;
`FIG. 6 is a block diagram of a logical interface mapping
`table in accordance with the present invention;
`FIG. 7 is a block diagram of a logical interface mapping
`table in accordance with the present invention;
`FIG. 8 is a block diagram of an exemplary on-volume table;
`FIG. 9 is a block diagram of an exemplary network envi
`ronment showing an IP volume connected to a switching
`network;
`FIG. 10 is a block diagram of an exemplary logical inter
`face mapping table in accordance with the present invention;
`FIG. 11 is a block diagram of an exemplary logical inter
`face mapping table in accordance with the present invention;
`and
`FIG. 12 is a block diagram of an exemplary network
`address to logical name mapping table in accordance with the
`present invention.
`
`DETAILED DESCRIPTION OF AN
`ILLUSTRATIVE EMBODIMENT
`
`A. Networks and File Servers
`
`20
`
`25
`
`30
`
`35
`
`40
`
`45
`
`50
`
`FIG. 1 is a schematic block diagram of an exemplary net
`work environment 100 in which the principles of the present
`invention are implemented. The network 100 is based around
`a local area network (LAN) interconnection 102. However, a
`wide area network (WAN), virtual private network (V PN)
`implementation (utiliZing communication links over the
`Internet, for example), or a combination of LAN, WAN and
`VPN implementations can be established. For the purposes of
`this description, the term “LAN” should be taken broadly to
`55
`include any acceptable networking architecture. The LAN
`102 interconnects various clients based upon personal com
`puters (PCs) 104, servers 106 and a network cache 107. Other
`network based devices such as X-terminals, workstations and
`the like can also be connected to the LAN. Also intercon
`nected to the LAN may be a switch/router 108 that provides a
`gateway to the well-known Internet 109, thereby enabling
`various networked devices to transmit and receive internet
`based information, including e-mail, web content, and the
`like.
`In addition exemplary ?le servers (or “?lers”) 110 and 112
`(Filer 1 and Filer 2, respectively) are connected to the LAN.
`
`60
`
`65
`
`6
`These ?lers, (described further below) are con?gured to con
`trol storage of, and access to, data in a set of interconnected
`storage volumes 122.
`Each of the devices attached to LAN include an appropri
`ate, conventional network interface arrangement (not shown)
`for communicating over the LAN using desired communica
`tion protocols, such as the well-known Transport Control
`Protocol/Internet Protocol (TCP/IP), User Datagram Proto
`col (U DP), Hypertext Transfer Protocol (HTTP) or Small
`Network Management Protocol (SNMP).
`While the invention is described herein in reference to a
`?ler or a cluster of ?lers, the teachings of this invention can be
`adapted to a variety of storage system architectures including,
`but not limited to, a network-attached storage environment, a
`storage area network and disk assembly directly-attached to a
`client/host computer. The term “storage system” should
`therefore be taken broadly to include such arrangements. It is
`expressly contemplated that the various processes, architec
`tures and procedures described herein can be implemented in
`hardware, ?rmware or software, consisting of a computer
`readable medium including program instructions that per
`form a series of steps. Moreover, while the invention is
`described herein in reference to ?le-based storage, it can be
`employed in other data storage schemas, such as block-based
`storage systems. Additionally, it should be noted that the term
`“volume” can also be de?ned to include a logical unit number
`(LUN) that designates one or more disks.
`Note that each ?ler can also be provided with a graphical
`user interface/console 150 so that instructions from an opera
`tor can be entered directly to the ?ler while generally bypass
`ing the LAN or other network.
`An exemplary ?le server, or ?ler, architecture is now
`described in further detail. FIG. 2 is a more-detailed sche
`matic block diagram of the exemplary ?le server 110 (Filer 1)
`implemented as a network storage appliance, such as the
`NetApp® ?ler available from Network Appliance, that can
`execute the above-described Data ONTAPTM software and is
`advantageously used with the present invention. Other ?lers
`can have similar construction (including exemplary Filer 2
`(112). By way of background, a network storage appliance is
`a special-purpose computer that provides ?le service relating
`to the organization of information on storage devices, such as
`disks. However, it will be understood by those skilled in the
`art that the inventive concepts described herein may apply to
`any type of ?ler whether implemented as a special-purpose or
`general-purpose computer, including a standalone computer.
`The ?ler 110 comprises a processor 222, a memory 224, a
`network adapter 226 and a storage adapter 228 intercon
`nected by a system bus 225. The ?ler 110 also includes a
`storage operating system 230 that implements a ?le system to
`logically organiZe the information as a hierarchical structure
`of directories and ?les on the disks.
`In the illustrative embodiment, the memory 224 may have
`storage locations that are addressable by the processor and
`adapters for storing software program code and data struc
`tures associated with the present invention. The processor and
`adapters may, in turn, comprise processing elements and/or
`logic circuitry con?gured to execute the software code and
`manipulate the data structures. The storage operating system
`230, portions of which are typically resident in memory and
`executed by the processing elements, functionally organiZes
`the ?ler 110 by, inter alia, invoking storage operations in
`support of a ?le service implemented by the ?ler. It will be
`apparent to those skilled in the art that other processing and
`memory means, including various computer readable media,
`may be used for storing and executing program instructions
`pertaining to the inventive technique described herein.
`
`LENOVO ET AL. EXHIBIT 1005
`Page 16 of 20
`
`

`
`US 7,707,263 B1
`
`7
`The network adapter 226 comprises the mechanical, elec
`trical and signaling circuitry needed to connect the ?ler 110 to
`a client 104 (including, for example, but not limited to, man
`agement station 140) (see FIG. 1) over the computer netWork
`(LAN 102), Which, as described generally above, can com
`prise a point-to-point connection or a shared medium, such as
`a local area netWork. A client (104, 140) can be a general
`purpose computer con?gured to execute applications includ
`ing ?le system protocols, such as the Common Internet File
`System (CIFS) protocol. Moreover, the client can interact
`With the ?ler 110 in accordance With a client/ server model of
`information delivery. That is, the client may request the ser
`vices of the ?ler, and the ?ler may return the results of the
`services requested by the client, by exchanging packets that
`conform to, e. g., the CIFS protocol format over the netWork
`102. The format of the CIFS protocol packet exchanged over
`the netWork is Well-knoWn and described in Common Internet
`File System (CIFS) Version: CIFS-Spec 0.9, Storage Net
`Working Industry Association (SNIA), Draft SNIA CIFS
`Documentation Work Group Work-in-Progress, Revision
`Date: Mar. 26, 2001 (hereinafter “CIFS speci?cation”),
`Which is hereby incorporated by reference as though fully set
`forth herein.
`The storage adapter 228 cooperates With the storage oper
`ating system 230 executing on the ?ler to access information
`requested by the client, Which information may be stored on
`a number of storage volumes 122. The storage adapter 228
`includes input/ output (I/O) interface circuitry that couples to
`the disks over an I/O interconnect arrangement, such as a
`conventional high-performance, Fibre Channel serial link
`topology. The information is retrieved by the storage adapter
`228 and, if necessary, processed by the processor 222 (or the
`adapter 228 itself) prior to being forWarded over the system
`bus 225 to the netWork adapter 226, Where the information is
`formatted into a packet and returned to the client 110.
`Notably, the exemplary ?ler 110 includes an NVRAM 260
`that provides faulttolerant backup of data, enabling the integ
`rity of ?ler transactions to survive a service interruption based
`upon a poWer failure, or other fault. The NVRAM 260 is
`typically made suf?ciently large to log a certain time-based
`chunk of transactions (for example, several seconds Worth).
`The NVRAM entry may be constructed in parallel With
`execution of the corresponding request, once it is determined
`that a request Will be successfully performed but it must be
`completed (as must any copying to mirror NVRAM of the
`partner in a cluster con?guration) before the result of the
`request is returned to the reque