CN112578994A - Cluster type storage system for expanding hard disk expansion unit - Google Patents
Cluster type storage system for expanding hard disk expansion unit Download PDFInfo
- Publication number
- CN112578994A CN112578994A CN201910934232.7A CN201910934232A CN112578994A CN 112578994 A CN112578994 A CN 112578994A CN 201910934232 A CN201910934232 A CN 201910934232A CN 112578994 A CN112578994 A CN 112578994A
- Authority
- CN
- China
- Prior art keywords
- hard disk
- expansion unit
- disk expansion
- host
- directly connected
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0604—Improving or facilitating administration, e.g. storage management
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0602—Interfaces specially adapted for storage systems specifically adapted to achieve a particular effect
- G06F3/0604—Improving or facilitating administration, e.g. storage management
- G06F3/0607—Improving or facilitating administration, e.g. storage management by facilitating the process of upgrading existing storage systems, e.g. for improving compatibility between host and storage device
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0628—Interfaces specially adapted for storage systems making use of a particular technique
- G06F3/0629—Configuration or reconfiguration of storage systems
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F3/00—Input arrangements for transferring data to be processed into a form capable of being handled by the computer; Output arrangements for transferring data from processing unit to output unit, e.g. interface arrangements
- G06F3/06—Digital input from, or digital output to, record carriers, e.g. RAID, emulated record carriers or networked record carriers
- G06F3/0601—Interfaces specially adapted for storage systems
- G06F3/0668—Interfaces specially adapted for storage systems adopting a particular infrastructure
- G06F3/0671—In-line storage system
- G06F3/0683—Plurality of storage devices
- G06F3/0689—Disk arrays, e.g. RAID, JBOD
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention provides a cluster type storage system for expanding hard disk expansion units, which comprises a host and a plurality of hard disk expansion units. The hard disk expansion units and the host form a tree structure connection with the host as a root node, each hard disk expansion unit is used for transmitting a device type indicating that the hard disk expansion unit is a hard disk expansion interface to the parent node and transmitting the type request to at least one corresponding child node after receiving a type request which is originated from a father node and is related to the device type, and each hard disk expansion unit transmits a device type indicating that the hard disk expansion unit is not a hard disk expansion interface to the father node after receiving a type request which is originated from a grander node and is related to the device type of the hard disk expansion unit, so that the host and each hard disk expansion unit can only identify the hard disk expansion unit directly connected with the host.
Description
[ technical field ] A method for producing a semiconductor device
The present invention relates to a cluster storage system, and more particularly, to a cluster storage system for expanding hard disk expansion units.
[ background of the invention ]
The conventional clustered storage system includes a host and at least one storage device connected to the host. The host is provided with a Linux operating system to operate the host and the storage device. Each storage device is, for example, a just a Bunch of disks (JBOD), and includes at least one hard disk expansion unit (expander), a plurality of hard disks electrically connected to the at least one hard disk expansion unit and each having a unique address.
The existing simple disk binding disk server uses hard disk expansion units to manage hard disks, the number of the hard disks that each hard disk expansion unit can connect to is limited by the number of the physical connection Ports (PHY) owned by the hard disk expansion unit, and the number of the hard disks that the simple disk binding disk server can install is jointly limited. In order to connect the raid disk server to more hard disks, the number of hard disks connected to the raid disk server can be increased by connecting a plurality of hard disk expansion units in series. However, in the architecture in which a plurality of hard disk expansion units are connected in series, not only the complexity of the host for managing the hard disk expansion units and the hard disks is increased, but also the original management architecture needs to be modified, and a solution is actually necessary to be provided.
[ summary of the invention ]
The present invention provides a cluster storage system for expanding hard disk expansion units, which reduces the complexity of the host managing the hard disk expansion units and the hard disks under the architecture of connecting a plurality of hard disk expansion units in series.
To solve the above technical problem, a cluster storage system for expanding hard disk expansion units comprises a host and a plurality of hard disk expansion units.
The hard disk expansion units and the host form a tree structure connection with the host as a root node, for each hard disk expansion unit, the hard disk expansion unit is used for transmitting a device type indicating that the hard disk expansion unit is a hard disk expansion interface to the father node and transmitting the type request to at least one corresponding child node after receiving a type request which is from a father node and is related to the device type, and after the hard disk expansion unit receives a type request which is from a grandfather node and is related to the device type of the hard disk expansion unit, the hard disk expansion unit transmits a device type indicating that the hard disk expansion unit is not a hard disk expansion interface to the grandfather node, so that the host and each hard disk expansion unit can only identify the hard disk expansion unit directly connected with the host and hide the hard disk expansion unit not directly connected with the host.
Another technical problem to be solved by the present invention is to provide a clustered storage system for expanding hard disk expansion units, which reduces the complexity of a host for managing the hard disk expansion units and hard disks in a structure in which a plurality of hard disk expansion units are connected in series.
Therefore, the cluster storage system for expanding hard disk expansion units of the invention comprises a host and a plurality of hard disk expansion units.
The hard disk expansion units and the host form a tree structure connection with the host as a root node, for the main hard disk expansion unit directly connected with the host, the main hard disk expansion unit is used for transmitting a device type indicating that the main hard disk expansion unit is a hard disk expansion interface to the host and transmitting the type request to at least one corresponding child node after receiving a type request which is from the host and related to the device type, for each other hard disk expansion unit which is not directly connected with the host, the other hard disk expansion unit is used for judging whether the type request is sent by the root node or not after receiving the type request related to the device type, when the other hard disk expansion unit judges that the type request is not sent by the root node, the other hard disk expansion unit transmits the type indicating that the device type is the hard disk expansion interface to be returned to the node which is sent by the source, and transmitting the type request to at least one corresponding child node, when the other hard disk expansion units judge that the type request is sent by the root node, the other hard disk expansion units transmit a type indicating that the type request is not a device type of a hard disk expansion interface to reply to the root node, so that the host can only identify the main hard disk expansion unit directly connected with the host and hide the other hard disk expansion units not directly connected with the host.
Compared with the prior art, the cluster storage system for expanding the hard disk expansion units has the advantages that for the hard disk expansion units which are not directly connected with the host, the host can only identify the hard disk expansion units which are directly connected with the host and hide the hard disk expansion units which are not directly connected with the host by transmitting the device type indicating that the hard disk expansion units are not hard disk expansion interfaces to the host, so that the complexity of the host for managing the hard disk expansion units and the hard disks is reduced, and the original management framework is not required to be modified.
[ description of the drawings ]
FIG. 1 is a block diagram illustrating a cluster storage system for expanding hard disk expansion units according to a first embodiment of the present invention.
[ detailed description ] embodiments
Referring to fig. 1, a first embodiment of a clustered storage system for expanding hard disk expansion units of the present invention includes a host 1, a plurality of hard disk expansion units 2a to 2e, and a plurality of hard disks 3a to 3 i. The hard disk expansion units 2 a-2 e and the host 1 form a tree structure connection (see the dotted line surrounding of fig. 1) with the host 1 as the root node, and each hard disk 3a/3b/3c/3d/3e/3f/3g/3h/3i is electrically connected to one of the hard disk expansion units 2 a-2 e.
For each hard disk expansion unit 2a/2b/2c/2d/2e, after receiving a type request related to the device type from a parent node directly connected to the hard disk expansion unit 2a/2b/2c/2d/2e, the hard disk expansion unit 2a/2b/2c/2d/2e is configured to transmit a device type indicating that the device type is a hard disk expansion interface back to the parent node, wherein the device type back further includes a unique address corresponding to the hard disk expansion unit and transmits the type request to at least one child node corresponding to the device type, and the hard disk expansion unit 2a/2b/2c/2d/2e receives a request related to the hard disk expansion unit 2a/2b/2c/2d 2e indirectly connected to a grandparent node and having a type request related to its own device type, the hard disk expansion unit 2a/2b/2c/2d/2e transmits a device type indicating that it is not a hard disk expansion interface back to the grandparent node of the hard disk expansion unit 2a/2b/2c/2d/2 e. It should be noted that each type request includes a unique Address, such as SAS Address, of a node originating the type request, so that each hard disk expansion unit 2a/2b/2c/2d/2e can identify an originating end of the corresponding type request according to the unique Address after receiving the type request. In this embodiment, the hdd expansion unit 2a/2b/2c/2d/2e indicates that it is not a hdd expansion interface by sending a device type response indicating that it does not have a device (no device), and the device type response does not include a unique address of the corresponding hdd expansion unit, but is not limited thereto. In addition, each hard disk expansion unit 2a/2b/2c/2d/2e stores a unique address of a parent node directly connected to the hard disk expansion unit 2a/2b/2c/2d/2e, thereby determining whether the received type request is from the parent node directly connected thereto.
It should be noted that when the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e scan the devices directly connected under it (i.e. their child nodes), it will determine whether to continue to scan the devices indirectly connected under it (i.e. their grandchild nodes) according to the device type of the devices directly connected under it, and when the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e identify the device type of its child node as the hdd expansion interface, the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e will continue to scan the device type of the grandchild node of the host 1/hdd expansion unit 2a/2b/2c/2d/2e (i.e., continuing to forward the type request to the grandchild node of the host 1/the HDD expansion units 2a/2b/2c/2d/2e via the child nodes of the host 1/the HDD expansion units 2a/2b/2c/2d/2e, when the host 1 and each HDD expansion unit 2a/2b/2c/2d/2e determines that the device type of the child node is not the HDD expansion interface, the host 1 and each HDD expansion unit 2a/2b/2c/2d/2e will stop scanning the device type of the grandchild node (i.e., stop forwarding the type request to the grandchild node via the child node). It should be noted that the direct connection does not refer to a direct connection of a physical location or a circuit, but indicates a direct connection relationship between the host 1 and each of the nodes of the hard disk expansion units 2a, 2b, 2c, 2d, and 2e, and there may be electronic components such as a hard disk, a backplane, a resistor, and the like, which are not considered by the nodes.
Taking the tree structure shown in fig. 1 as an example, the host 1 sends a type request to its corresponding child node (i.e., the hdd expansion unit 2a), and the hdd expansion unit 2a sends a device type indicating that it is an hdd expansion interface back to the host 1. Since the device type responded by the host 1 is the hard disk expansion interface, the host 1 continues to forward the type request to its grandchild node (i.e., the hard disk expansion units 2b, 2c) via its child node, the hard disk expansion unit 2a forwards the type request from the host 1 to the two child nodes (i.e., the hard disk expansion units 2b, 2c) corresponding to the hard disk expansion unit 2a, after the hard disk expansion unit 2b receives a type request of the device type from a grandparent node of the hard disk expansion unit 2b and related to the hard disk expansion unit 2b itself, the hard disk expansion unit 2b transmits a type request indicating that it is not a hard disk expansion interface to the grandparent node of the hard disk expansion unit 2b, similarly, after the hard disk expansion unit 2c receives a type request of the device type from the grandparent node and related to the hard disk expansion unit 2c itself, the hard disk expansion unit 2c transmits a device type indicating that it is not a hard disk expansion interface back to the grandparent node of the hard disk expansion unit 2 c. Since the host 1 receives the reply of the hdd expansion unit 2c and determines that the device type of the grandchild node is not the hdd expansion interface, the host 1 stops forwarding the type request to the child nodes (i.e., the hdd expansion units 2d and 2e) under the grandchild node through the child nodes and the grandchild nodes thereof, so that the host 1 only determines the device type of the hdd expansion unit 2a as the hdd expansion interface, thereby hiding the hdd expansion units 2b, 2c, 2d, and 2 e. The hard disk expansion unit 2a transmits a type request to its corresponding child node (i.e., the hard disk expansion units 2b, 2c), the hard disk expansion unit 2b transmits a device type indicating that it is a hard disk expansion interface back to its parent node, i.e., the hard disk expansion unit 2a, and similarly, the hard disk expansion unit 2c transmits a device type indicating that it is a hard disk expansion interface back to its parent node, i.e., the hard disk expansion unit 2 a. Since the hard disk expansion unit 2a receives the reply of its child node to determine that the device type of its child node (i.e., the hard disk expansion unit 2b) is the hard disk expansion interface, the hard disk expansion unit 2a continues to forward the type request to its grandchild node (i.e., the hard disk expansion units 2d, 2e) through its child node, that is, the hard disk expansion unit 2b forwards the type request from the hard disk expansion unit 2a to its corresponding two child nodes, after the hard disk expansion unit 2d receives the type request of the device type related to the hard disk expansion unit 2d itself from the grandparent node of the hard disk expansion unit 2d, the hard disk expansion unit 2d transmits a device type indicating that it is not the hard disk expansion interface back to the grandparent node of the hard disk expansion unit 2d, and similarly, after the hard disk expansion unit 2e receives a device type request related to the device type of the hard disk expansion unit 2d itself from the grandparent node of the hard disk expansion unit 2e and related to itself After requesting the type of the device type, the hard disk expansion unit 2e transmits a device type response indicating that it is not a hard disk expansion interface to its grandparent node. Since the hard disk expansion unit 2a receives the reply of the grandchild node thereof and determines that the device type of the grandchild node is not the hard disk expansion interface, the hard disk expansion unit 2a stops forwarding the type request to other child nodes below the grandchild node thereof through the child nodes and the grandchild nodes thereof (in the example of fig. 1, there are no other child nodes below the hard disk expansion units 2d and 2 e). Since the hard disk expansion unit 2a determines that the device type of the other child node (i.e., the hard disk expansion unit 2c) is the hard disk expansion interface, the hard disk expansion unit 2a continues to forward the type request to the grandchild node of the hard disk expansion unit 2a through the other child node, but since the hard disk expansion unit 2c has no other child node below it but is a leaf node, the hard disk expansion unit 2c will not forward the type request from the hard disk expansion unit 2a, and thus for the hard disk expansion unit 2a, it will only determine the hard disk expansion units 2b, 2c as the hard disk expansion interface, so as to achieve the purpose of hiding the hard disk expansion units 2d, 2 e. The hdd expansion unit 2b transmits a type request to its corresponding child node (i.e., the hdd expansion units 2d and 2e), the hdd expansion unit 2d transmits a device type response indicating that it is an hdd expansion interface to the hdd expansion unit 2b, and similarly, the hdd expansion unit 2e transmits a device type response indicating that it is an hdd expansion interface to the hdd expansion unit 2 b. Since the hard disk expansion unit 2b receives the reply of its child node and recognizes the device type of its child node as a hard disk expansion interface, the hard disk expansion unit 2b continues to forward the type request to the grandchild node through its child node (i.e., the hard disk expansion unit 2d), but since the hard disk expansion unit 2d has no child node below it as a leaf node, the hard disk expansion unit 2d does not forward the type request originated from the hard disk expansion unit 2b any more, and similarly, the hard disk expansion unit 2b continues to forward the type request to the grandchild node through its other child node (i.e., the hard disk expansion unit 2e), but since the hard disk expansion unit 2e has no child node below it as a leaf node, the hard disk expansion unit 2e does not forward the type request originated from the hard disk expansion unit 2b any more, so for the hard disk expansion unit 2b, it will only determine the hard disk expansion units 2d, 2e as hard disk expansion interfaces.
The hard disk expansion unit 2a/2b/2c/2d/2e only replies the device type of the host 1 or the hard disk expansion unit directly connected with the host 1 or the hard disk expansion unit with the hard disk expansion interface, but not replies the device type of the host 1 or the hard disk expansion unit directly connected with the host 1 or the hard disk expansion unit with the hard disk expansion interface, so that the host 1 and each hard disk expansion unit 2a/2b/2c/2d/2e can only identify the hard disk expansion unit directly connected with the host by the received reply and hide the hard disk expansion unit not directly connected with the host, and therefore, the host 1 regards the hard disk expansion unit 2b, 2c, 2d and 2e not directly connected with the host 1 as nonexistent. It should be noted that the host 1 only uses the hard disk expansion unit with the device type as the hard disk expansion interface as the managed node under the management architecture, so that when the host 1 scans the nodes of the management architecture of the clustered storage system, for the elements that do not belong to the hard disk expansion interface (i.e. the nodes that reply the host 1 and whose device type is not the hard disk expansion interface), the host 1 regards the nodes that reply the host 1 and which are not the hard disk expansion interface as the non-existent nodes in the management architecture.
For each hard disk expansion unit 2a/2b/2c/2d/2e, the hard disk expansion unit 2a/2b/2c/2d/2e is further configured to determine whether there is at least one hard disk expansion unit directly connected to the hard disk expansion unit 2a/2b/2c/2d/2e other than its parent node after receiving a hard disk data request from its parent node and related to the hard disk under it. It should be noted that, for each hard disk expansion unit 2a/2b/2c/2d/2e, the hard disk under the hard disk expansion unit 2a/2b/2c/2d/2e refers to the hard disk connected in series with the hard disk expansion unit, including direct connection and indirect connection, but not between the hard disk expansion unit and the host, and does not include other hard disks connected in parallel with the hard disk expansion unit. When the hard disk expansion unit 2a/2b/2c/2d/2e determines that the at least one directly connected hard disk expansion unit exists, the hard disk expansion unit 2a/2b/2c/2d/2e transmits the data request to the at least one directly connected hard disk expansion unit, after the hard disk expansion unit 2a/2b/2c/2d/2e receives the data reply of each directly connected hard disk expansion unit, each data reply contains the unique Address of the hard disk under the data reply, such as SAS Address, the hard disk expansion unit 2a/2b/2c/2d/2e generates and transmits a unique Address containing the hard disk under each directly connected hard disk expansion unit and the unique Address of the hard disk directly connected with the hard disk expansion unit 2a/2b/2c/2d/2e The address data is returned to the father node of the hard disk expansion unit 2a/2b/2c/2d/2e, and the hard disk expansion unit 2a/2b/2c/2d/2e also stores the corresponding relation between each directly connected hard disk expansion unit and the unique address of the hard disk under the directly connected hard disk expansion unit. When the hard disk expansion unit 2a/2b/2c/2d/2e determines that there is no at least one directly connected hard disk expansion unit, the hard disk expansion unit 2a/2b/2c/2d/2e generates and transmits a data including a unique address of a hard disk directly connected to the hard disk expansion unit 2a/2b/2c/2d/2e back to its parent node.
Taking the tree structure shown in fig. 1 as an example, for the hard disk expansion unit 2d, when the hard disk expansion unit 2d receives a hard disk data request from the hard disk expansion unit 2b and related to the hard disk under it, the hard disk expansion unit 2d determines that there is no hard disk expansion unit directly connected to it except its parent node (i.e., outside the hard disk expansion unit 2b), so the hard disk expansion unit 2d generates and transmits a data containing the unique address of the hard disk 3b, 3c directly connected to the hard disk expansion unit 2d back to the hard disk expansion unit 2 b. Similarly, for the hard disk expansion unit 2e, when the hard disk expansion unit 2e receives a hard disk data request from the hard disk expansion unit 2b and related to the hard disk under it, the hard disk expansion unit 2e determines that there is no at least one hard disk expansion unit directly connected thereto other than its parent node (i.e., outside the hard disk expansion unit 2b), so the hard disk expansion unit 2e generates and transmits a data reply to the hard disk expansion unit 2b containing the unique address of the hard disk 3d, 3e, 3f directly connected to the hard disk expansion unit 2 e. For the hard disk expansion unit 2b, when the hard disk expansion unit 2b receives a hard disk data request from the hard disk expansion unit 2a and related to the hard disk under it, the hard disk expansion unit 2b determines that there is at least one directly connected hard disk expansion unit 2d, 2e in addition to its parent node (i.e., in addition to the hard disk expansion unit 2a), so that the hard disk expansion unit 2b transmits the data request to the at least one directly connected hard disk expansion unit 2d, 2e, after the hard disk expansion unit 2b receives the data reply of each directly connected hard disk expansion unit 2d, 2e, the hard disk expansion unit 2b generates and transmits a unique address including the hard disk 3b, 3c, 3d, 3e, 3f under each directly connected hard disk expansion unit 2d, 2e, and the data of the unique address of the hard disk 3g directly connected with the hard disk expansion unit 2b is replied to the hard disk expansion unit 2 a.
It should be noted that the host 1 is also stored in the management structure of the clustered storage system, and replies the host 1 with a unique address corresponding to the hard disk expansion unit 2a whose device type is the hard disk expansion interface. Unlike the prior art, all the hard disk expansion units 2a, 2b, 2c, 2d, 2e receiving the type request of the host 1 in the management structure of the clustered storage system reply the host 1 with its device type being a hard disk expansion interface, so that the number of the hard disk expansion units that can be managed by the host 1 in the management structure of the clustered storage system is limited by the number of physical connection Ports (PHYs) of the host 1. By adopting the technology proposed by the present application, in the management structure of the cluster storage system, the host 1 does not store the unique address corresponding to the hard disk expansion units 2b, 2c, 2d, 2e whose device types are not hard disk expansion interfaces, so that the physical connection ports of the host 1 are not occupied, and the management structure is not directly managed by the host 1, and the number of all the hard disk expansion units 2a, 2b, 2c, 2d, 2e in the management structure of the cluster storage system is not limited by the number of physical connection Ports (PHY) of the host 1, as long as the number of the hard disk expansion units 2a directly connected with the host 1 does not exceed the number of the physical connection ports of the host 1.
Referring to fig. 1, a hardware architecture of a second embodiment of the clustered storage system for expanding hard disk expansion units according to the present invention is the same as the first embodiment, and the difference is that a mechanism for hiding the hard disk expansion units is different. How the hard disk expansion unit is hidden in the second embodiment will be described below.
For the main hdd expansion unit 2a directly connected to the host 1, the main hdd expansion unit 2a is configured to, after receiving a type request related to a device type from the host 1, transmit a device type indicating that the device type is hdd expansion interface back to the host 1 and forward the type request to at least one corresponding child node (i.e., other hdd expansion units 2b and 2c), for each other hdd expansion unit 2b/2c/2d/2e not directly connected to the host 1, the other hdd expansion unit 2b/2c/2d/2e is configured to, after receiving a type request related to a device type, determine whether the type request originates from the root node (i.e., host 1), and when the other hdd expansion unit 2b/2c/2d/2e determines that the type request does not originate from the root node (i.e., host 1) When the root node sends out, the other hard disk expansion unit 2b/2c/2d/2e transmits a type indicating the type of the device which is the hard disk expansion interface to the node which sends out the type request, and transmits the type request to at least one corresponding child node, when the other hard disk expansion unit 2b/2c/2d/2e judges that the type request is sent out by the root node, the other hard disk expansion unit 2b/2c/2d/2e transmits a type indicating the type of the device which is not the hard disk expansion interface to the root node, so that the host 1 can only identify the main hard disk expansion unit 2a which is directly connected with the host 1, and hide the other hard disk expansion units 2b, 2c, 2d, 2e which are not directly connected with the host. Each of the hdd expansion units 2a/2b/2c/2d/2e stores a unique address of the root node (i.e., the host 1), so as to determine whether the received type request originates from the root node. Each type request includes a unique Address, such as SAS Address, of a node originating the type request, so that each hard disk expansion unit 2a/2b/2c/2d/2e can identify an originating end of the corresponding type request according to the unique Address after receiving the type request. In this embodiment, the other hdd expansion units 2b/2c/2d/2e indicate that they are not hdl interfaces by sending a device type response indicating that they do not have a device (no device) to the root node, but is not limited thereto.
It should be noted that, like the first embodiment, when the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e scan the devices (i.e. child nodes) connected thereunder, it will be determined whether to continue scanning the devices (i.e. grandchild nodes) connected thereunder according to the device types of the devices connected thereunder, when the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e identify the device type of the child node as the hdd expansion interface, the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e will continue scanning the device type of the grandchild node (i.e. continue forwarding the type request to the grandchild node through the child node), and when the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e identify the child node thereof When the device type of the point is not the hdd expansion interface, the host 1 and each hdd expansion unit 2a/2b/2c/2d/2e stop scanning the device types of their grandchild nodes (i.e., stop forwarding type requests to grandchild nodes via child nodes), so that for the host 1, the host 1 will transmit the type request to the main hard disk expansion unit 2a and forward the type request to the other hard disk expansion units 2b, 2c through the main hard disk expansion unit 2a, however, since the other hard disk expansion units 2b and 2c both transmit a type reply indicating that they are not the device type of the hard disk expansion interface to the host 1, the host 1 therefore stops forwarding type requests to the child nodes under the grandchild node (i.e., the other hard disk expansion units 2d, 2e) through the child nodes and grandchild node. For the main hdd expansion unit 2a, the main hdd expansion unit 2a will transmit a type request related to the device type to at least one corresponding sub-node, and for each other hdd expansion unit 2b/2c/2d/2e directly or indirectly connected to the main hdd expansion unit 2a, the other hdd expansion unit 2b/2c/2d/2e is configured to transmit the type indicating the device type as the hdd expansion interface back to the main hdd expansion unit 2a and transmit the type request to at least one corresponding sub-node after determining that the type request is originated from the main hdd expansion unit 2 a. Taking the tree structure shown in fig. 1 as an example, the main hdd expansion unit 2a transmits the type request to its corresponding child node (i.e., the other hdd expansion units 2b and 2c), the other hdd expansion unit 2b transmits a device type indicating that it is an hdd expansion interface back to the main hdd expansion unit 2a, and similarly, the other hdd expansion unit 2c transmits a device type indicating that it is an hdd expansion interface back to the main hdd expansion unit 2 a. Since the main expansion unit 2a determines that the device type of its child node (i.e., the other expansion unit 2b) is the hdd interface, the main expansion unit 2a continues to forward the type request to the grandchild node (i.e., the other expansion units 2d, 2e) via the child node, the other expansion unit 2b forwards the type request from the main expansion unit 2a to its corresponding two child nodes, after the other expansion unit 2d receives the type request related to its own device type from the main expansion unit 2a, the other expansion unit 2d transmits a device type indicating that it is the hdd interface back to the main expansion unit 2a, and similarly, after the other expansion unit 2e receives the type request related to its own device type from the main expansion unit 2a, the other hard disk expansion unit 2e sends a device type indicating that it is a hard disk expansion interface back to the main hard disk expansion unit 2a, so that the main hard disk expansion unit 2a continues to forward the type request to the child node below the grandchild node through the child node and the grandchild node (in the example of fig. 1, there is just no child node below the other hard disk expansion units 2d, 2e), so that the other hard disk expansion units 2d, 2e do not forward the type request from the main hard disk expansion unit 2a any more. Since the main hdd expansion unit 2a determines that the device type of the other child node (i.e., the hdd expansion unit 2c) is the hdd expansion interface, the main hdd expansion unit 2a continues to forward the type request to the grandchild node through the other child node, but since the hdd expansion unit 2c is a leaf node and there is no child node therebelow, the hdd expansion unit 2c will not forward the type request from the main hdd expansion unit 2 a.
When the main hard disk expansion unit 2a directly connected with the host 1 receives a hard disk data request from the host 1 and related to the hard disk under it, the main hard disk expansion unit 2a sends another data request to all other hard disk expansion units 2b, 2c, 2d, 2e directly or indirectly connected with it, after the main hard disk expansion unit 2a receives the data reply of each other hard disk expansion unit 2b/2c/2d/2e, each data reply contains the unique Address of the hard disk under it, such as SAS Address, the main hard disk expansion unit 2a generates and transmits a unique Address containing the hard disk 3g/3h, 3i/3b, 3c/3d, 3e, 3f under each other hard disk expansion unit 2b/2c/2d/2e, and the data of the unique address of the hard disk 3a directly connected with the main hard disk expansion unit 2a is returned to the father node, and the main hard disk expansion unit 2a also stores the corresponding relation of the unique address of each other hard disk expansion unit 2b/2c/2d/2e and the hard disk 3g/3h, 3i/3b, 3c/3d, 3e and 3f under the other hard disk expansion units in the management structure of the clustered storage system.
In summary, the cluster storage system for expanding hard disk expansion units of the present invention transmits a device type indicating that the hard disk expansion unit is not a hard disk expansion interface to the host 1 in response to the hard disk expansion unit not directly connected to the host 1, so that the host 1 can only identify the hard disk expansion unit directly connected to the host and hide the hard disk expansion unit not directly connected to the host, thereby reducing the complexity of the host 1 in managing the hard disk expansion interface and the hard disk, and without modifying the original management structure, thereby achieving the objective of the present invention. It should be noted that the node of the management architecture of the clustered storage system stored in the host 1 only includes a unique address of the main hard disk expansion unit 2a directly connected to the host 1 (in other embodiments of the present invention, the number of the main hard disk expansion units 2a is not limited to only one, as long as the hard disk expansion units directly connected to the host 1 are all the main hard disk expansion units 2a, and the number of the main hard disk expansion units 2a is required to meet the limitation of the number of physical connection Ports (PHY) of the host 1, that is, the number of the main hard disk expansion units 2a is not greater than the number of physical connection ports of the host 1.
The above description is only for the specific embodiments of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily conceive of the changes or substitutions within the technical scope of the present invention, and all the changes or substitutions should be covered within the scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (9)
1. A clustered storage system for expanding hard disk expansion units, comprising:
a host computer; and
a plurality of hard disk expansion units, wherein the hard disk expansion units and the host form a tree structure connection with the host as a root node, the hard disk expansion unit is used for receiving a type request which originates from a father node and is related to the device type, the hard disk expansion unit transmits a device type indicating that the hard disk expansion unit is a hard disk expansion interface to the father node and transmits the type request to at least one corresponding child node, after the hard disk expansion unit receives a type request originating from a grandparent node and relating to its own device type, the hard disk expansion unit transmits a device type response indicating that the hard disk expansion unit is not a hard disk expansion interface to the grandparent node, the host and each hard disk expansion unit can only identify the hard disk expansion unit directly connected with the host and hide the hard disk expansion unit not directly connected with the host.
2. The clustered storage system for expanding hard disk expansion units as claimed in claim 1, further comprising:
a plurality of hard disks, each hard disk being electrically connected to one of the hard disk expansion units;
wherein, for each hard disk expansion unit, the hard disk expansion unit is also used for judging whether at least one hard disk expansion unit directly connected with the hard disk expansion unit exists except the father node of the hard disk expansion unit after receiving a hard disk data request which is from the father node of the hard disk expansion unit and is related to the hard disk under the father node of the hard disk expansion unit, when the hard disk expansion unit judges that the at least one directly connected hard disk expansion unit exists, the hard disk expansion unit transmits the data request to the at least one directly connected hard disk expansion unit, after the hard disk expansion unit receives the data reply of each directly connected hard disk expansion unit, each data reply comprises the unique address of the hard disk under the hard disk expansion unit, the hard disk expansion unit generates and transmits the data comprising the unique address of the hard disk under the directly connected hard disk expansion unit and the unique address of the hard disk directly connected with the hard disk expansion unit to return to the father node of the hard disk expansion unit, when the hard disk expansion unit judges that the at least one directly connected hard disk expansion unit does not exist, the hard disk expansion unit generates and transmits data containing the unique address of the hard disk directly connected with the hard disk expansion unit to reply to the father node of the hard disk expansion unit.
3. The clustered storage system for expanding hard disk expansion units as claimed in claim 2, wherein for each hard disk expansion unit, when the hard disk expansion unit determines that there is at least one directly connected hard disk expansion unit, after the hard disk expansion unit receives the data reply of each directly connected hard disk expansion unit, the hard disk expansion unit further stores the corresponding relationship between each directly connected hard disk expansion unit and the unique address of the hard disk under the directly connected hard disk expansion unit.
4. The clustered storage system for expanding hard disk expansion units of claim 1 wherein the type request includes a unique address of a node from which the type request originates.
5. A clustered storage system for expanding hard disk expansion units, comprising:
a host computer; and
a plurality of hard disk expansion units, the hard disk expansion units and the host form a tree structure connection with the host as a root node, for the main hard disk expansion unit directly connected with the host, the main hard disk expansion unit is used for transmitting a device type which is a hard disk expansion interface back to the host and transmitting the type request to at least one sub-node corresponding to the host after receiving a type request which is from the host and related to the device type, for each other hard disk expansion unit which is not directly connected with the host, the other hard disk expansion unit is used for judging whether the type request is sent by the root node or not after receiving a type request related to the device type, when the other hard disk expansion unit judges that the type request is not sent by the root node, the other hard disk expansion unit transmits a type which is the hard disk expansion interface back to the origin to send the source When the other hard disk expansion unit judges that the type request is originated from the root node, the other hard disk expansion unit transmits a type indicating that the type is not a device type of a hard disk expansion interface to the root node, so that the host can only identify a main hard disk expansion unit directly connected with the host and hide other hard disk expansion units not directly connected with the host.
6. The clustered storage system as claimed in claim 5, wherein the main expansion unit directly connected to the host transmits a type request related to a device type to the corresponding at least one child node, and for each of the other expansion units directly or indirectly connected to the main expansion unit, the other expansion unit transmits a type indicating that the type of the device is a hdd interface back to the main expansion unit and transmits the type request to the corresponding at least one child node.
7. The clustered storage system for expanding hard disk expansion units as claimed in claim 5, further comprising:
a plurality of hard disks, each hard disk being electrically connected to one of the hard disk expansion units;
when the main hard disk expansion unit directly connected with the host receives a hard disk data request which comes from the host and is related to the hard disk under the host, the main hard disk expansion unit sends another data request to all other hard disk expansion units directly or indirectly connected with the host, after the main hard disk expansion unit receives the data reply of each other hard disk expansion unit, each data reply comprises the unique address of the hard disk under the host, the main hard disk expansion unit generates and transmits a data reply comprising the unique address of the hard disk under the other hard disk expansion unit and the unique address of the hard disk directly connected with the main hard disk expansion unit to the father node of the main hard disk expansion unit.
8. The clustered storage system as claimed in claim 7, wherein after the main expansion unit receives the data reply from each other hard disk expansion unit, the hard disk expansion unit further stores the corresponding relationship between each other hard disk expansion unit and the unique address of the hard disk under the other hard disk expansion unit.
9. The clustered storage system for expanding hard disk expansion units as claimed in claim 5, wherein the type request includes a unique address of a node originating the type request.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910934232.7A CN112578994A (en) | 2019-09-29 | 2019-09-29 | Cluster type storage system for expanding hard disk expansion unit |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910934232.7A CN112578994A (en) | 2019-09-29 | 2019-09-29 | Cluster type storage system for expanding hard disk expansion unit |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112578994A true CN112578994A (en) | 2021-03-30 |
Family
ID=75111288
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910934232.7A Pending CN112578994A (en) | 2019-09-29 | 2019-09-29 | Cluster type storage system for expanding hard disk expansion unit |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112578994A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5365589A (en) * | 1992-02-07 | 1994-11-15 | Gutowitz Howard A | Method and apparatus for encryption, decryption and authentication using dynamical systems |
| US20060041672A1 (en) * | 2004-08-18 | 2006-02-23 | Day Brian A | Systems and methods for initiator mode connection management in SAS connections |
| US8296513B1 (en) * | 2009-08-21 | 2012-10-23 | Pmc-Sierra, Inc. | Standby disk spin up control in disk array system using expander device |
| CN103176913A (en) * | 2011-12-26 | 2013-06-26 | 英业达股份有限公司 | Hard disk dynamic mapping method and server for applying same |
| US20140068633A1 (en) * | 2012-08-29 | 2014-03-06 | Kyocera Document Solutions Inc. | Multi processing system |
| US20140173165A1 (en) * | 2012-12-13 | 2014-06-19 | Lsi Corporation | Expander for Loop Architectures |
| US20160147478A1 (en) * | 2014-11-21 | 2016-05-26 | Fujitsu Limited | System, method and relay device |
| TWI566168B (en) * | 2015-11-05 | 2017-01-11 | 神雲科技股份有限公司 | Routing method for cluster storage system |
| CN107102695A (en) * | 2016-02-22 | 2017-08-29 | 佛山市顺德区顺达电脑厂有限公司 | The method of the installation position of the abnormal hard disk of judgement for cluster type stocking system |
| CN206684730U (en) * | 2017-04-20 | 2017-11-28 | 郑州云海信息技术有限公司 | The system that a kind of PCIE of storage server extends direct-connected hard disk |
-
2019
- 2019-09-29 CN CN201910934232.7A patent/CN112578994A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US5365589A (en) * | 1992-02-07 | 1994-11-15 | Gutowitz Howard A | Method and apparatus for encryption, decryption and authentication using dynamical systems |
| US20060041672A1 (en) * | 2004-08-18 | 2006-02-23 | Day Brian A | Systems and methods for initiator mode connection management in SAS connections |
| US8296513B1 (en) * | 2009-08-21 | 2012-10-23 | Pmc-Sierra, Inc. | Standby disk spin up control in disk array system using expander device |
| CN103176913A (en) * | 2011-12-26 | 2013-06-26 | 英业达股份有限公司 | Hard disk dynamic mapping method and server for applying same |
| US20140068633A1 (en) * | 2012-08-29 | 2014-03-06 | Kyocera Document Solutions Inc. | Multi processing system |
| US20140173165A1 (en) * | 2012-12-13 | 2014-06-19 | Lsi Corporation | Expander for Loop Architectures |
| US20160147478A1 (en) * | 2014-11-21 | 2016-05-26 | Fujitsu Limited | System, method and relay device |
| TWI566168B (en) * | 2015-11-05 | 2017-01-11 | 神雲科技股份有限公司 | Routing method for cluster storage system |
| CN107102695A (en) * | 2016-02-22 | 2017-08-29 | 佛山市顺德区顺达电脑厂有限公司 | The method of the installation position of the abnormal hard disk of judgement for cluster type stocking system |
| CN206684730U (en) * | 2017-04-20 | 2017-11-28 | 郑州云海信息技术有限公司 | The system that a kind of PCIE of storage server extends direct-connected hard disk |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103268318B (en) | A kind of distributed key value database system of strong consistency and reading/writing method thereof | |
| US7380074B2 (en) | Selecting storage clusters to use to access storage | |
| US7818408B1 (en) | Automated virtual interface failover in a mass storage cluster | |
| US8095639B2 (en) | Monitoring-target-apparatus management system, management server, and monitoring-target-apparatus management method | |
| US20210075665A1 (en) | Implementing switchover operations between computing nodes | |
| WO2018107772A1 (en) | Method, device and apparatus for processing write request | |
| US20030182592A1 (en) | Failure detection and failure handling in cluster controller networks | |
| CN112130748B (en) | Data access method, network card and server | |
| CN102571452A (en) | Multi-node management method and system | |
| JP5686034B2 (en) | Cluster system, synchronization control method, server device, and synchronization control program | |
| US20120284435A1 (en) | Zone group manager virtual phy | |
| US20040162921A1 (en) | SCSI enclosure services | |
| US20120239789A1 (en) | Storage device, controller, and address management method | |
| EP2597818A1 (en) | Cluster management system and method | |
| CN106936931B (en) | Method, related equipment and system for realizing distributed lock | |
| US11463376B2 (en) | Resource distribution method and apparatus in Internet of Things, device, and storage medium | |
| US20100332532A1 (en) | Distributed directory environment using clustered ldap servers | |
| CN104580425A (en) | Client data synchronization method and system | |
| US8065401B2 (en) | Systems and methods for frame ordering in wide port SAS connections | |
| JP4653490B2 (en) | Clustering system and method having interconnections | |
| CN112578994A (en) | Cluster type storage system for expanding hard disk expansion unit | |
| US9218313B2 (en) | Shared-bandwidth multiple target remote copy | |
| CN112805964A (en) | Method and apparatus for ensuring continuous device operational reliability in cloud degradation mode | |
| US20150135004A1 (en) | Data allocation method and information processing system | |
| TWI720528B (en) | Clustered storage system for expanding expander |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |