System and methods to manage wide storage area network

[0001] This invention relates generally to networks, and more particularly, but not exclusively, provides a system and methods for managing a wide storage area network.

[0002] A Storage Area Network (SAN) is a network of storage disks and may also include switches, servers, and disk subsystems. A benefit of SANs is storage consolidation, which enables the sharing of a single or multiple large disk subsystems across a large number of servers and/or applications. Another benefit of SANs is that a plurality of SANs may be interconnected with bridges, such as a WDM device, and located at different geographic locations to form Wide SANs. For example, a SAN in San Francisco may be communicatively coupled to a SAN in San Diego.

[0003] SANs and Wide SANs are managed using SAN management software. An important feature of conventional SAN management software is automatic discovery of all SAN devices, such as servers, switches, and disk subsystems, in a SAN. SAN management software also stores a topology of the discovered SAN devices. Therefore, a new system and method for managing SANs is needed.

[0004] The present invention provides a system for correctly recognizing SAN topology. Conventional SAN management software cannot recognize geographic distance between devices, such as bridges. Accordingly, in a situation in which a bridge is between two geographically separated SANs, an IT administrator may inadvertently set up a zone among the two geographically separated SANs because the management software doesn't recognize the geographic difference between the two SANs. This may lead to lower performance and complicated management problems.

[0005] To solve the aforementioned limitation, the system of the present invention comprises a SAN manager engine, a topology repository, a discovery list and a configuration table. The SAN manager engine communicates with management agents of devices of the SAN to retrieve configuration data from each agent. The management agents in switches store, in addition to configuration information, distance and/or response time between switch ports and target devices. The switch management agents can know distance and/or response time via IT manager manual input; measuring reply time from sending a frame from a switch to a target device; and/or receiving location data from a Global Positioning System (GPS) receiver or similar device. The SAN manager also checks configurations and warns a user if a bad configuration is discovered.

[0006] The topology repository includes data from all SAN devices as collected by the SAN manager from manager agents. The SAN manager can also generate a topology tables based on the data in the repository. The discovery list provides information on all SAN devices. The SAN manager uses the discovery list to contact management agents of the devices. The configuration history table provides historical configuration data in the SAN.

[0007] The present invention further provides methods for managing a wire storage area network. The methods comprise a topology discovery method and a configuration check method. The topology discovery method comprises discovering devices in a SAN per a discovery list; gathering data from management agents in the discovered devices; comparing the gathered data with data stored in a topology repository; if gathered data is different from stored data, copying the stored data to a configuration history table and storing the gathered data in the topology repository; and outputting the results in a graphical and/or table format. The configuration check method comprises examining the topology repository for zoning between switches separated by a long distance and displaying a warning if a zone has been set up with switches separated by a long distance.

[0008] Accordingly, the system and method advantageously provides for improved management of wide SANs.

[0024] The following description is provided to enable any person skilled in the art to make and use the invention, and is provided in the context of a particular application and its requirements. Various modifications to the embodiments will be readily apparent to those skilled in the art, and the principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles, features and teachings disclosed herein.

[0025] FIG. 1 is a block diagram illustrating a SAN in which a bridge has no management agent and uses a transparent protocol. The SAN includes a local SAN in San Diego (“Local SAN SD”) and a local SAN in San Francisco (“Local SAN SF”). The Local SAN SD comprises a disk subsystem D 11000; a Switch C 12000; a server 13000; a management server 14000; management network 15000; and a bridge E 23000. The management network 15000 is used for communicating with Management Server 14000 and all the SAN Devices to exchange the configuration tables. Network 15000 may use any type of protocol, such as Fibre Channel, SCSI, FC over IP, etc.

[0026] Disk subsystem D 11000 comprises a management agent 11100; volumes 11300(v1 and v2); and disk ports 11200(d1 and d2). Management agent 11100 manages the configuration of the disk subsystem 11000. Agent 11100 also has two tables including a Disk Port Table (11110) and Volume Path Table (11120). These tables are in further detail below. Disk ports 11200 include data connection ports using a fibre channel protocol between servers 13000 and disk subsystem 11000. In an alternative embodiment of the invention, each disk port 11200 uses a protocol such as SCSI, FC over IP, or any other protocol. Volume 11300(v1 and v2) includes volumes that are exported to the servers 13000. The volume 11300 may comprise a single physical disk drive, or multiple disk drives where the data is striped and managed by a disk array controller, or any other physical structure.

[0027] Switch 12000 is an interconnect device between servers 13000 and disk subsystem 11000. In an embodiment of the invention, there is at least one switch that supports a zoning function in a SAN. The switch 12000 may use any communication protocol. The switch 12000 comprises a management agent 12100 and switch ports 12200(s1-s6). Agent 12100 manages configuration of switch 12000. The agent 12100 includes two tables including a Port Link Table 12110 and a Zoning Table 12120 as will be discussed in further detail below. Switch Port 12200 is a data interconnection port between servers 13000 and disk subsystem 11000. There is at least one switch port 12200 in a switch, and any connection protocol, such as Fibre Channel, SCSI, FC over IP, etc. can be used.

[0028] Servers 13000(a and b) are application servers for use in a SAN. The servers 13000 each comprise management agents 13100(a and b); and host ports 13200(a1, a2 and b1). The agents 13100 are located in servers 13000 and manage the configuration of the servers 13000. In addition, these agents 13100 each have a Host Port Table 13110, which will be discussed in further detail below. Host Ports 13200(a1, a2, b1) are data connection ports between servers 13000 and disk subsystem 11000. There is at least one host port in each server, and each host port 13200 uses a connection protocol, such as Fibre Channel, SCSI, FC over IP, or any other protocol. In an embodiment, Server 13000a has 2 Fibre Channel host ports, 13200-a1 and 13200-a2 and Server 13000b has 1 Fibre Channel host port 13200-b1.

[0029] Management Server 14000 is for management use in the SAN. The management server 14000 comprises a SAN manager 14100 that communicates with all the Management Agents, such as agents 11100, 12100, 13100 and 15100 (in FIG. 1-1), in a SAN to get the configuration information via Management Network 15000. In addition, SAN manager 14100 communicates with SAN manager 22100 to share data between their respective topology repositories 14120 and 22120. The SAN manager 14100 also includes a topology repository 14120, a discovery list 14110 and a configuration table 14140. The repository 14120 stores all table data from the SAN devices such as Servers, Switches and Disk Subsystems (from both Local SAN SD and Local SAN SF). The discovery list, 14110, configuration table 14140, and the topology repository 14120 will be discussed in further detail below.

[0030] Management Network 15000 is a network for communicating with Management Servers 14100 and all the SAN Devices to exchange configuration tables. The connection protocols may include Fibre Channel, SCSI, FC over IP, or any other protocol.

[0031] Bridge 15000 is an interconnect device between different protocol networks. The bridge 15000 comprises bridge ports 15200(e1 and e2) that can use any protocol, such as Fibre Channel, SCSI, FC over IP, etc.

[0032] Interconnect 20000 is used for long distance communication between bridges 23000 and 16000. Interconnect 20000 can use any protocol. SAN managers 14100 and 22100 can both recognize the distance between bridges 23000 and 16000 and represent the interconnection as a “cloud” or by simply displaying “Long Distance.”

[0033] Local SAN SF is substantially similar to Local SAN SD and includes a Server C 18000 having a management agent 18100 (having a host port table 18110) and a host port 18200; a disk subsystem H 19000 having a management agent 19100 (having a disk port table 19110 and volume path table 19120), a volume 19300, and a disk port 19200; a bridge 16000 having ports 16200-f1 and 16200-f2; a management server 22000 having a SAN manager 22100 (including a discovery list 22110, topology repository 22120, and configuration table 22140); and Switch G 17000 comprising a management agent 17120 (having a port link table 17110 and zoning table 17120 ) and ports 17200(g1-g4).

[0034] The components of Local SAN SF are substantially similar to the components of Local SAN SD and therefore operate in a substantially similar way. Via management network 21000, management server 22000 is coupled to server 18000, switch 17000 and components of Local SAN SD. In addition, bridge 16000 is coupled to switch port 17200-g3 of switch 17000 via port 16200-f2; disk subsystem 19000 is coupled to switch port 17200-g2 via disk port 19200-h1; and server 18000 is coupled to port switch 17200-g1 via host port 18200-c.

[0035] FIG. 1-1 is a block diagram of a wide SAN in which the bridges have management agents and use a non-transparent protocol function. The wide SAN of FIG. 1-1 is substantially similar to the wide SAN of FIG. 1 except that bridge 23000 and 16000 each include a management agent 15000 and 16000 respectively. Each management agent 15000 and 16000 each include a port link table 15110 and 16110 respectively. The management agents 23000 and 16000 each manage the respective configurations of bridges 23000 and 16000 respectively. Management agents 16000 and 23000 and their respective port link tables 16110 and 15110 will be discussed in further detail below.

[0036] FIGS. 2 and 2-1 are diagrams of disk port tables 11110 and 19110 of disk subsystem 11000 and 19000 respectively. These tables 11110 and 19110 provide information regarding disk ports in a disk subsystem. These tables each have two columns including a disk port ID (210), which is the name for each disk port; and WWN (220), which is a unique identifier for each disk port. When using a fibre channel protocol, WWN (220) stores a World Wide Name for each disk port.

[0037] FIGS. 3 and 4 are diagrams illustrating the volume path tables 11120 and 19120 in disk subsystems 11000 and 19000 respectively. These tables 11120 and 19120 provide information of the data I/O path from disk ports to volumes, referred to as “Volume Path” or simply “Path”. These tables each have five columns including: Path ID (310), which is the name for the path; Disk Port ID (320), which is the Disk Port ID attached to the path; Volume ID (330), which is the name for the volume attached to the path; SCSI ID (340), which is the SCSI ID attached to the path (IT administrators apply this value to the server as the SCSI ID); and SCSI LUN (350), which is the SCSI LUN (Logical Unit Number) attached to the path (IT administrators apply this value to the server as the SCSI LUN).

[0038] FIGS. 5, 5-1, 5-2, and 5-3 are diagrams of port link tables 12110 and 17110 of Switches 12000 and 17000 respectively. FIGS. 5 and 5-1 correspond to the wide SAN of FIG. 1 while FIGS. 5-2 and 5-3 correspond to the wide SAN of FIG. 1-1. These port link tables 12110 and 17110 provide information relating to the interconnect relationship between servers and disk subsystems via switches, which are referred to as links. These tables 12110 and 17110 each has five columns, as follows:

[0039]

Switch Port ID (510):	This is a name for each switch port.
Switch Port WWN (520):	This is a unique identifier of each
	switch port. In a Fibre Channel
	protocol, a World Wide Name (WWN)
	is used.
Link WWN (530):	This shows the target device WWN
	connected to this switch port. This
	may be the host port's, other switch
	port's, or disk port's WWN.
Dist. (540):	This shows the distance between a
	switch port and a target device. The
	distance information may be
	quantitative as in “300 miles” or
	qualitative as in “far”. In this
	embodiment, we use a qualitative
	value. Distance can be determined by
	reading attributes that were manually
	set in the switches by an IT
	administrator or by reading attributes
	set via a GPS receiver coupled to a
	switch.
Res. time (550):	This shows the response time to send
	and receive a data frame, such as a
	login frame, between a switch port and
	target device. The response time
	information may be quantitative as in
	“100 msec” or qualitative as in “long”.
	In this embodiment, we use a
	qualitative value.

[0040] In an embodiment of the invention, either Distance (540) or Response Time (550) is used.

[0041] FIG. 6 and FIG. 6-1 are diagrams of zoning tables 12120 and 17120 in FC Switch C 12000 and FC Switch G 17000 respectively. These tables 12120 and 17120 provide zoning configuration data in switches 12000 and 17000 respectively. Tables 12120 and 17120 include a list of switch ports that make up each zone. Two switch ports in the same zone can communicate with each other, but if switch ports are located in different zones they are unable to communicate with each other. The tables 12120 and 17120 each have two columns including a Zone ID (610), which include names for each zone, and a switch port ID list (620), which lists all switch ports in each zone.

[0042] FIG. 7a, FIG. 7b, and FIG. 7c are diagrams showing host port tables 13110a, 13110b and 18110c in server 13000-a, server 13000-b, server 18000-c respectively. Tables 13110a, 13110b and 18110c provide information relating to host ports in servers 13000-a, 13000-b and 18000-c respectively. Each table has three columns, including a Host Port ID (710), which is a name for each host port; WWN (720), which is a unique identifier of each host port (when using Fibre Channel protocol, WWN stores the World Wide Name); and SCSI ID (730), which is the SCSI ID assigned to the specified Host Port by IT administrators (this ID is uniquely assigned in each server).

[0043] FIG. 8a, FIG. 8b, and FIG. 8c are diagrams showing Logical Unit Number (LUN) Binding tables 13120-a, 13120-b and 18120-c in servers 1300-a, 13000-b and 18000-c respectively. These tables 13120-a, 13120-b and 18120-c provide the information relating to the data I/O path from the host port to the SCSI Logical Unit (LU), which is referred to as “LUN Binding” or simply “Binding”. These tables 13120-a, 13120-b and 18120-c each have four columns as follows:

[0044]

Binding ID (810):	This is the name for the binding.
SCSI ID (820):	This is the SCSI ID attached to the
	binding.
LUN (830):	This is the SCSI LUN attached to the
	binding.
Inquiry Information (840):	This is the information given by the
	LUN when servers issue a SCSI
	INQUIRY Command to the LUN.
	This information may contain
	vendor name, product name, and
	volume ID of the LUN.

[0045] FIG. 9 and FIG. 9-1 are diagrams of port link tables 15110 and 16110 in bridges 23000 and 16000 (of FIG. 1 using a transparent protocol) respectively. FIG. 9-2 and FIG. 9-3 are diagrams of port link tables 15110 and 16110 in bridges 23000 and 16000 (of FIG. 1-1 using a non-transparent protocol) respectively. These table 15110 and 16110 provide information relating to the interconnect relationship between different protocol networks via bridges, which are referred to as “links”. These tables 15110 and 16110 each have three columns including a Bridge Port ID (910), which is the nickname for each bridge port; Bridge Port Address (920), which is a unique identifier of each bridge port (in a Fibre Channel protocol, we use the World Wide Name (WWN) as the unique identifier); and Link Address (930), which is the target device address connected to this bridge port. To determine the link address in the case of transparent protocol, frames can be snooped so as to read header information including source and destination address. To determine the link address in the case of a non-transparent protocol, the WWN of a SAN device to which the bridge is linked is retrieved.

[0046] FIG. 10 is a diagram illustrating discovery list 14110, which is identical to discovery list 22110 in management server 22000. FIG. 10 corresponds to the wide SAN of FIG. 1 having transparent bridges, while FIG. 10-1 corresponds to the wide SAN of FIG. 1-1 having non-transparent bridges. These tables 14110 and 22110 provide information on all the devices in a wide SAN. SAN Manager 14000 uses this table to get the configuration information from Management Agents in the SAN devices. These tables 14110 and 22110 each have six columns as follows:

[0047]

Local SAN ID (1060):	This identifies the SAN that the
	device belongs to.
Discovery ID (1010):	This is the name for the target SAN
	device to be discovered.
Device Type (1020):	This column shows the device type
	of the entry. SAN Manager 14000
	uses this column to see what kind of
	tables to expect to receive.
Device Information (1030):	This is detailed information of the
	target SAN device.
IP Address (1040):	This is the communication point of
	the Management Agent in a target
	device. In this embodiment, we
	suppose that SAN Manager 14000
	communicates with all the SAN
	devices via TCP/IP.
Position (1050):	This shows the latitude longitude or
	general area where the device is
	located. In an embodiment, either
	general area information or latitude
	longitude information is required to
	recognize the distance between each
	SAN device if there is no distance
	information or response time
	information provided by port link
	tables 15110 or 16110 in switches
	23000 or 16000 respectively.

[0048] FIG. 11-1 is a diagram showing a topology table 14130 in management server 14000 in a wide SAN having bridges using a transparent protocol. FIG. 11-2 is a diagram showing a topology table 14130, which is identical to topology table 22130, in a wide SAN having bridges using a non-transparent protocol. SAN Manager (14100, 22100) creates these tables 14130 and 22130 that provide the topology of the I/O communication in a wide SAN. SAN Manager (14100, 22100) makes this table by merging the Disk Port Table (11110, 19110), Volume Path Table (11120, 19120), Port Link Table (12110, 17110), Host Port Table (13110, 18110), Discovery List (14110, 22110) and LUN Binding Table (13130, 18130). These tables each have three columns as follows:

[0049]

Server (1110):	This column shows the server information in a
	topology entry. This column has three sub-
	columns, which provide detail information of the
	server. These sub-columns are Server Name
	(1111), Binding ID (1112), and Host Port ID
	(1113).
Interconnect (1120):	This column shows the switch and bridge
	information in a topology entry. This column
	has four sub-columns, which provide detail
	information on the switch and bridge. These
	sub-columns are Switch Name (1121) and
	Switch Port ID (1122) on one side, and Switch
	Name (1123) and Switch Port ID (1124) on the
	other side. Further, there is Distance (1129) or
	Area (1129) information in FIG. 11-1. Further,
	there is Bridge Name (1125) and Bridge Port ID
	(1126) on one side, and Bridge Name (1127) and
	Bridge Port ID (1128) on the other side in FIG.
	11-2.
Storage (1130):	This column shows the disk subsystem
	information in a topology entry. This column has
	3 sub-columns, which provides detail
	information of the disk subsystem. These sub-
	columns are Storage Name (1131), Disk Port ID
	(1132), and Volume ID (1133).

[0050] FIG. 12 is an example of configuration table 14140, which is identical to table 22140, in server 14000. These tables 14140 and 22140 provide the configuration in a SAN. SAN Manager (14100, 22100) stores the information of any SAN devices' configurations in tables 14140 and 22140. These tables are used when SAN Manager (14100, 22100) analyzes the bad configuration caused by the configuration of a zone that includes a long distance interconnect. The configuration table 14140 can also store prior configurations. These tables each have three columns as follows:

[0051]

Device Name (1210):	This column shows the device name.
Configuration type (1220):	This column shows the
	configuration type of devices.
Configuration detail (1230):	This column shows the detailed
	information of the configuration.
	This column has 2 sub-columns. ID
	(1231), which shows ID of the
	configuration and Detail (1232),
	which shows the detail of the
	configuration.

[0052] FIGS. 13A and 13B are flowcharts illustrating methods 1300 and 1305. SAN Manager (14100 and/or 22100) can execute methods 1300 and 1305 concurrently or one at a time. First, SAN manager (14100 and/or 22100) requests (1310) data from all management agents in all SAN components. For example, SAN manager (14100 and/or 22100) can request port link table 17110 and zoning table 17120 from agent 17100 in switch 17000. Next, SAN manager (14100 and/or 22100) receives (1320) the data from the agents. The SAN manager (14100 and/or 22100) compares (1330) the received data with data already stored in the SAN manager's respective topology repository (14120 or 22120). If the data differs, the SAN manager (14100 and/or 22100) updates (1340) its associated repository (14120 or 22120) and stores (1340) old configuration data in configuration table 14140 and/or 22140. The SAN manager (14100 and/or 22100) then creates or updates (1350), if necessary, a topology table (14130 or 22130) based on the data in the repository (14120 or 22140). The SAN manager (14100 and/or 22100) then outputs the results in a graphical format showing the topology and/or in table format (e.g., displaying table 14130 or 22130). In a graphical format, long distance between bridges and/or other components can be shown as a cloud, by text indicating a long distance, or via other techniques. The method 1300 then end.

[0053] Method 1305 comprises for each configuration in Topology Repository 14120 or 22140, SAN Manager (14100 or 22100) checks (1370) for a bad configuration wherein there is a long distance connection between switches as indicated in the Configuration Table (14140, 22140), Topology table (14130, 22130) and/or Discovery list (14110, 22110). If there is a bad configuration, then the SAN Manager (14100 or 22100) outputs (1380) a warning regarding the bad configuration. The warning can include which zones and/or switches comprise the bad configuration as well as the switches locations. The method 1305 then ends.

[0054] The foregoing description of the preferred embodiments of the present invention is by way of example only, and other variations and modifications of the above-described embodiments and methods are possible in light of the foregoing teaching. For example, a GPS receiver for determining position can be replaced with a GLONASS receiver or other location-determining device. Although the network sites are being described as separate and distinct sites, one skilled in the art will recognize that these sites may be a part of an integral site, may each include portions of multiple sites, or may include combinations of single and multiple sites. Further, components of this invention may be implemented using a programmed general purpose digital computer, using application specific integrated circuits, or using a network of interconnected conventional components and circuits. Connections may be wired, wireless, modem, etc. The embodiments described herein are not intended to be exhaustive or limiting. The present invention is limited only by the following claims.