The present invention generally relates to computer storage and more specifically, to a system and method for policy-based storage provisioning.
As applications become more advanced and complicated, the allocation of storage for the applications becomes increasingly difficult. Storage area networks (SANs), for example, provide applications running on computer systems (servers, server clusters, PCs, and so on) with high storage capacity capability using a plurality of storage server systems coupled to the computer systems. Typically, SAN administrators manually allocate storage based upon the needs of an application running on a server or server cluster that will use disk space. Additionally, when there are any changes, such as in the characteristics of the application, the characteristics of the server or server cluster, the capability of the storage devices, the network topology, or the availability of storage in a data center, the administrator may need to revise earlier decisions on storage allocations and take necessary corrective action, such as migrating existing data from existing disks to new disks. The changing of earlier decisions of storage allocation may be complex, time-consuming, and error-prone, and may require application downtime.
Frequently, storage administrators do not have the necessary information and/or tools to process available information to take the appropriate actions. Further, even if the necessary information and tools are available to take the corrective actions, the effort and cost involved may deter the administrators from taking the actions. Thus, sub-optimal usage of storage capacity results and the performance of an application is subsequently degraded.
Another issue that SAN administrators face is the heterogeneity of storage configuration interfaces for each of the different types of storage devices that are deployed in a SAN. Each of the storage configuration interfaces have different capabilities and nuances that make it very difficult, if not impossible, for SAN administrators to configure storage in a consistent and uniform manner. Thus, deploying new storage devices becomes very difficult and may even be delayed or avoided even though the new storage device may be more suitable or cost effective for the applications that use the storage.
In one embodiment, an apparatus and method for provisioning and managing storage using storage provisioning policies is provided. Storage provisioning policies are created and are used to select and provision appropriate storage for applications.
A storage provisioning policy is created by specifying storage heuristics for storage attributes using storage heuristic metadata. Storage attributes characterize a storage device and storage heuristic metadata describe how to specify a storage heuristic. Using the storage heuristic metadata, storage heuristics are defined to express a rule or constraint as a function of a storage attribute. In addition, the storage provisioning policy may also specify mapping rules for exporting the storage to a consumer of the storage, such as the server or server cluster.
After storage provisioning policies are created, discovered data for storage attributes associated with the storage devices are compared to related storage heuristics for the storage attributes found in the storage profile. Then, one or more storage devices are selected for provisioning if the discovered attributes of the storage devices satisfy the storage heuristics in the storage profile. The storage that is provisioned may be any type of storage, such as storage units of storage devices, storage devices, and the like. In another embodiment, virtual media units may be created if no storage devices satisfy the storage heuristics in the storage profile. The selected or created storage units are then provisioned to the storage consumer using the mapping rules specified as part of the storage provisioning policy.
In one embodiment, a method for provisioning one or more storage devices using one or more storage provisioning policies and discovered storage attribute data from the one or more storage devices is provided. The one or more storage provisioning policies comprising one or more storage heuristics for one or more storage attributes. The method comprises: comparing the discovered storage attribute data associated with the one or more storage devices to the one or more storage heuristics for the storage attributes in the one or more storage provisioning policies; and if discovered storage attribute data associated with a storage device in the one or more storage devices satisfy the one or more storage heuristics for the one or more storage attributes, provisioning the storage device.
In another embodiment, a method for creating one or more storage provisioning policies is provided. The method comprises: defining one or more storage attributes; defining one or more storage heuristic metadata associated with the one or more storage attributes; and specifying one or more storage heuristics using the defined one or more storage heuristic metadata associated with the one or more defined storage attributes to create the storage provisioning policy, the storage provisioning policy usable to provision a storage device, wherein the provisioned storage device includes discoverable data that satisfies the storage heuristics for the storage attributes.
A further understanding of the nature and advantages of the invention herein may be realized by reference of the remaining portions in the specifications and the attached drawings.
The present application is related to other applications, filed on this date, Feb. 13, 2001, and assigned to the present assignee, that are directed to other aspects of storage servers and that claim priority to U.S. Provisional Application No. 60/268,694 above: U.S. Nonprovisional patent application Ser. No. 00/______, Attorney Docket No. 20949P-000200US entitled “Silicon Based Storage Virtualization Server”; U.S. Nonprovisional patent application Ser. No. 00/______, Attorney Docket No. 20949P-000300US entitled “Storage Virtualization and Storage Management to Provide Higher Level Storage Services”; U.S. Nonprovisional patent application Ser. No. 00/______, Attorney Docket No. 20949P-000500US entitled “Method and Apparatus for Identifying Storage Devices”; U.S. Nonprovisional patent application Ser. No. 00/______, Attorney Docket No. 20949P-000700US entitled “Virtual Data Center”; U.S. Nonprovisional patent application No. 00/______, Attorney Docket No. 20949P-000800US entitled “Failover Processing in a Storage System”; U.S. Nonprovisional patent application No. 00/______, Attorney Docket No. 20949P-000900US entitled “RAID at Wire Speed” and U.S. Nonprovisional patent application No. 00/______, Attorney Docket No. 20949P-001000US entitled “Multi-Level Hardware Based Device Security”. These applications are incorporated herein by reference for all purposes.
In one embodiment of the present invention, storage provisioning policies are created. The policies may be based on factors, such as Quality of Storage Service (QoSS) or the application requirements. Once a policy is created, the policy may be used to automatically provision storage, monitor storage, and take corrective actions. The policies may be used across a wide variety of storage devices allowing uniform deployment of storage provisioning policies in a heterogeneous storage environment.
The SAN 102 can be any type of computer network. It is referred to as a storage area network in the present application because that is its relevant function with respect to the embodiments of the present invention. In an embodiment of the present invention, the SAN 102 is a Fibre Channel network, the host computers 106 and the storage devices 102 are configured to communicate with a Fibre Channel network, and the storage server 100 is also configured to communicate with a Fibre Channel network. Thus, the storage server 100 can be easily added to an existing SAN.
The physical storage devices 104 include tape drives, disk arrays, JBODs (“just a bunch of disks”), or other types of data storage devices. The physical storage devices 104 can be connected directly to the host computers 106 via the SAN 102 or can be indirectly connected to the host computers 106 via the SAN 102 and the storage server 100. Management of storage virtualization is improved by using the storage server 100 to indirectly connect the storage devices 104 to the host computers 106.
The host computers 106 can be servers or stand-alone computers. The host computers 106 can be directly connected to the SAN 102 or indirectly connected via a switch, router, or other communication link.
The storage server 100 may include one or more storage processors 110. The storage processors 110 process the storage commands and data to be stored as information flows between the host computers 106 and the storage devices 104. One or more of the storage processors 110 may be included on each line card 112. The storage server 100 includes space for numerous line cards 112, so the capabilities of the storage server 100 can be modularly increased by adding more line cards 112 or more storage processors 110. Each storage processor 110 is associated with one or more ports of the storage server 100.
The storage server 100 may include one or more virtual server cards 114. The virtual server cards control the operation of the storage server 100 and control the line cards 112, which perform the actual work of transferring commands and data.
The switch fabric 116 connects the storage processors 110. The switch fabric switches information received at one port to another port of the storage server 100. For example, when a host computer 106 wants to read data stored on the storage area network 102, its request is processed by the storage processor 110 associated with the port associated with that host computer 106. That storage processor 110 is referred to as the upstream storage processor 110. The upstream storage processor 110 communicates with a downstream storage processor 110 associated with the port associated with the storage device 104 storing the data to be read, via the switch fabric 116. Then the switch fabric 116 transfers the data read from the storage device to the host computer 106, via the downstream and upstream storage processors 110.
VSX 100 allocates storage to host computers 106 based upon the needs of applications running on host computers 106. The process of configuring and exporting storage to host computers 106 is called storage provisioning. The storage provisioned may be any type of storage, such as storage units of storage devices, storage devices, and the like. The term storage device will be used hereinafter, and it will be understood that storage devices may represent any type of storage that may be provisioned.
In one embodiment, VSX 100 facilitates the process of storage provisioning by allowing a user to specify policies based upon different criteria such as, for example, the data access pattern of the application, availability of storage in the SAN, the reliability and data availability requirements for the data to be stored, etc. Additionally, once a Storage Provisioning policy has been established by the user, other users may use it to provision storage to servers. It will be understood that a user may be an administrator, application, software agent working on behalf of the application, and the like. VSX 100 facilitates the discovery of supported Storage Attributes, and the storage heuristics and storage heuristic metadata needed to create storage-provisioning policies based upon the supported attributes. This allows users to dynamically create and adapt the policies based upon the supported attributes.
In one embodiment, storage provisioning policies are created using storage attributes, storage heuristics, and storage heuristic metadata. Also, other actions, such as the mapping rules and corrective actions, may be specified in a storage provisioning policy. In creating a storage provisioning policy, an application's storage requirements are defined. Storage heuristics embodying the requirements are specified for storage attributes and stored in a storage profile. The specification of storage heuristics for storage attributes in the storage profile creates a storage provisioning policy that is used to allocate storage to a server.
Storage mapping rules may be specified and stored in the storage profile. However, in one embodiment, the storage mapping rules are stored in a storage mapping profile. It will be understood that the elements of the storage profile and storage mapping profile may be stored anywhere that is accessible to VSX 100. The mapping rules are used during the provisioning of storage to the storage consumer, such as the server or server cluster running the application for which the storage is required. Also, the storage mapping profile is used to handle changes to the storage consumer. For example, if a new server is added to a server cluster, for which storage has been provisioned, the storage mapping profile is used to automatically export the provisioned storage to the new server.
Multiple storage profiles may be created for different needs. For example, a storage profile may be created for allocating storage for databases and another for allocating storage for streaming applications. Also, policies may specify actions for the different needs, such as provisioning a certain minimum amount of available storage capacity in the SAN that satisfy a profile or provisioning a certain minimum amount of available storage capacity in the SAN that satisfy a profile that is available for immediate use.
A storage attribute is a property of storage that is used to characterize the storage. For example, an attribute may characterize a storage's performance, reliability, availability, cost, etc. The storage attributes of storage devices need to be discovered by software in VSX 100 in order to characterize the storage device. For example, storage attributes are discovered by systems and methods described in Attorney Docket No. 20949P-000500US entitled, “Method and Apparatus for Identifying Storage Devices,”. In addition, as described hereinafter, VSX 100 may modify or create storage attributes to change the storage characteristic in order to provision storage. For example, changing the SCSI Mode Pages or creating a mirror set using several disks the characteristic of the storage may be changed.
Table 1 illustrates examples of storage attributes, corresponding descriptions of the storage attributes, and units in which the storage attributes may be measured according to one embodiment. Also, it will be understood that this list is not exhaustive and that a person skilled in the art will appreciate other storage attributes.
A set of possible storage attributes are defined and stored in VSX 100. In another embodiment, the defined set may be stored remotely but be accessible to VSX 100.
In one embodiment, the valid domain may be a range, list, or a list that is localized to the users language based on the local settings. A localized list is a list that has different discrete values. Each of these values has a different description in different languages. An example of a localized list is Multi-portedness—this list has two values single-ported and multi-ported in English and will have a different set of two values in French, Spanish, Japanese, Korean, etc.
Additionally, storage heuristic metadata specifies the data type and the unit used to mathematically express the storage heuristic for the storage attribute. Additionally, a constraint type may be defined, such as value, range, list, Boolean, etc. One example of storage heuristic metadata is specifying a range of 0 to 10,000 MB/second for the storage attribute maximum transfer rate and a constraint of Boolean. Thus, the metadata specifies that discovered data from storage devices for the attribute may be in the range of 0 to 10,000 MB/second and a constraint for the heuristic is a Boolean.
Storage heuristic metadata provides the building block for defining storage heuristics. In one embodiment, a user may discover all defined storage attributes by querying VSX 100 for storage attribute metadata. Once receiving the metadata for all attributes, the user specifies heuristics for desired storage attributes using the metadata.
The storage heuristic is specified from the defined storage heuristic metadata. For example, a storage heuristic metadata for the storage attribute maximum transfer rate has the domain range of 0 to 10,000 MB/second and a constraint of a range. A storage heuristic is specified from the domain. For example, a storage heuristic might be a range for the transfer rate of 200 to 3000. Thus, the rule is access speeds between 200 MB/second and 3000 MB/second.
The storage mapping rules may also be specified for the storage provisioning policy. The mapping rules specify rules for exporting the storage to a consumer of the storage, such as the server or server cluster. The mapping rules include policies on LUN ( SCSI Logical Unit Number), the access rights to the data on the storage, etc.
Table 2 illustrates examples of storage mapping rules. Also, it will be understood that this list is not exhaustive and that a person skilled in the art will appreciate other storage mapping rules.
Policies may also be specified for taking actions if certain conditions occur. For example, a policy for handling changes to a storage profile may specify an action to automatically migrate existing storage allocated to storage consumers using the storage profile to new storage that satisfies the new storage profile requirements. Thus, VSX 100 monitors the storage profile and when changes to the profile are detected, VSX 100 automatically migrates existing storage to new storage that meets the changed profile. This action is specified in the storage profile.
Also, corrective actions may be included in a policy. For example, VSX 100 monitors the storage heuristics specified in the storage profile used to allocate storage to a server. If any of the constraints of the storage heuristics are violated or changed, VSX 100 takes a corrective action that is specified in the storage profile. Examples of actions include but are not limited to alerting the SAN administrator, automatically migrating existing storage to storage that meets the QoSS requirements, etc. Additionally, changes to the storage consumer (the server or server cluster for which the storage has been provisioned) are handled by using the mapping rules specified in the storage mapping profile.
In step S600, storage attributes that are supported by VSX 100 are selected and defined. These attributes are the supported attributes that define what characteristics will be used to provision storage.
In step S602, storage heuristic metadata is defined. The storage heuristic metadata describe the supported storage attributes and are used to specify the rules for creating storage heuristics.
In step S604, storage heuristics are specified for the storage attributes using the domain and units of the storage heuristic metadata. The heuristics may be based on application requirements or QoSS standards.
Once the storage heuristics are defined, other requirements or actions may be defined (step S606). For example, as described above, corrective actions, storage mapping rules, or additional policies may be specified.
In step S608, a storage profile and a storage mapping profile are created that define the storage provisioning policy. The storage profile is created using the defined storage heuristics and other requirements and actions. The storage mapping profile is created by specifying the mapping rules for the storage. The storage profile and storage mapping profile may be stored in VSX 100 or be stored remotely but be accessible by VSX 100.
In one embodiment, an administrator creates a storage provisioning policy. In another embodiment, applications or software agents acting on behalf of the application create the storage provisioning policies. Policies may also be created dynamically from a policy that is being enforced. In this case, changes in characteristics of the storage, application, or profile may cause policies to be dynamically created. For example, storage heuristic metadata may have a range of 1-100,000 MB/second. An action may be created indicating that if discovered data is greater than 100,000 MB/second, an action should be taken by VSX 100 to create a new policy using new storage heuristics for the new data. Additionally, storage heuristic metadata for the storage attribute may be changed because of the new discovered data.
In step S700, VSX 100 queries for and receives discovered storage attribute data for supported storage attributes. Additionally, the VSX may receive all discovered storage attribute data and filter the data for only the supported attributes in the profile.
In step S702, the discovered storage attribute data from the storage devices are compared to the storage heuristics in a storage profile. In step S704, VSX 100 determines if any storage devices satisfy the constraints or rules of the storage heuristics.
If a storage device does include data that satisfy the storage heuristics, VSX 100 may automatically provision the storage device to a host computer 106 (step S706). In one embodiment, the provisioning is done using storage mapping rules. Additionally, if more than one storage device satisfies the heuristics, VSX 100 may provision all the devices or a policy in the storage profile may specify how many storage devices are provisioned.
If no storage device includes data that satisfy the storage heuristics, in step S708, VSX 100 determines if a virtual storage unit may be created to satisfy the storage heuristics of the storage profile. If not, an error condition is created and the user is informed that the provisioning policy cannot be satisfied (S712). If a virtual storage unit is capable of being created, VSX 100 creates one or more virtual storage units (virtual Media Units) that satisfy the storage heuristics and the virtual storage unit is provisioned to the server (S710). A virtual Media Unit is a unit of storage that is constructed from the physical storage (disks) available in the network. A virtual Media Unit has the same behavior as physical units of storage and may be used by host computer 106 but has characteristics that are usually different from that the underlying physical units of storage used to create the virtual Media Unit. Examples of virtual Media Units include but are not restricted to concatenations, mirror sets, and RAID of different types. In another embodiment, VSX 100 may also try to change the characteristics of the physical storage devices to meet the constraints, if necessary and possible, by, for example, setting mode sense pages if it allows the disk to meet the specified constraints.
In step S714, VSX 100 monitors the discovered storage attribute data for the provisioned storage unit and determines if the data meet the storage heuristics specified in the storage provisioning policy. In step S714 changes to the characteristics of the storage consumers are also monitored.
In step S716, VSX 100 determines if the heuristics are being satisfied and that there are no changes to the characteristics of the storage consumer. The monitoring ensures that QoSS is being met and that the characteristics of the storage consumer have not changed. 1621 If the heuristics are being met and there are no changes to the characteristics of the storage consumer, the process reiterates back to step S714 to continue monitoring discovered storage attribute data.
If the heuristics are not being met or there are changes to the characteristics of the storage consumer, in step S718, VSX 100 determines if corrective action may be taken.
If corrective actions are needed, in step S724, it is determined whether the corrective actions are with respect to changes in the characteristics of the storage consumer or with respect to the QoSS requirements.
If corrective actions are needed with respect to the QoSS requirements not being met, in step S720, it is determined if data needs to be migrated to a new storage device. If so, the process proceeds to step S708, where it is determined if a virtual storage unit may be created.
If data does not need to be migrated, in step S722, VSX 100 performs the corrective actions to meet the storage heuristics specified in the storage provisioning policy.
If corrective actions are needed with respect to changes in the characteristics of the storage consumer, in step S726, the VSX 100 performs the corrective actions to accommodate changes to the characteristics of the storage consumer using the rules specified in the storage mapping profile of storage provisioning policy.
Embodiments of the present invention provide an apparatus and method to create storage provisioning policies. The creation of the policies is facilitated using storage attributes, storage heuristics, and storage metadata. Using supported attributes and storage heuristic metadata, storage heuristics are specified for a storage profile. Storage meeting the constraints of the heuristics is then provisioned. Thus, policies may be created using QoSS or applications requirements and storage is automatically provisioned using the storage profile.
The above description is illustrative but not restrictive. Many variations of the invention will become apparent to those skilled in the art upon review of the disclosure. The scope of the invention should, therefore, be determined not with reference to the above description, but instead should be determined with reference to the pending claims along with their full scope or equivalents.
This application claims priority to U.S. Provisional Application No. 60/268,694, filed Feb. 13, 2001,entitled “Virtual Storage Systems” which is incorporated herein by reference.
Number | Date | Country | |
---|---|---|---|
60268694 | Feb 2001 | US |
Number | Date | Country | |
---|---|---|---|
Parent | 10077482 | Feb 2002 | US |
Child | 10928081 | Aug 2004 | US |