Customizing data management for a distributed filesystem

• US 9,678,981 B1
• Filed: 02/15/2013
• Issued: 06/13/2017
• Est. Priority Date: 05/03/2010
• Status: Active Grant

• Current Assignee: Panzura, Inc.
• Original Assignee: Panzura, Inc.
• Inventors: Taylor, John Richard, Chou, Randy Yen-pang, Davis, Andrew P.
• Primary Examiner(s): Rahman, Sm
• Application Number:
  US13/769,185
• Time in Patent Office: 1,579 Days
• Field of Search: 709219
• CPC Class Codes: G06F 11/1435 using file system or storage system metadata
• CPC Class Codes: G06F 11/2089 Redundant storage control functionality
• CPC Class Codes: G06F 11/2094 Redundant storage or storage space G06F11/2056 takes precedence
• CPC Class Codes: G06F 11/3006 where the computing system is distributed, e.g. networked systems, clusters, multiprocessor systems multiprogramming arrangements G06F9/46; allocation of resources G06F9/50
• CPC Class Codes: G06F 11/3433 for load management allocation of a server based on load conditions G06F9/505; load rebalancing G06F9/5083; redistributing the load in a network by a load balancer H04L67/1029
• CPC Class Codes: G06F 16/122 using management policies backup systems G06F11/1446; file migration policies for HSM systems G06F16/185
• CPC Class Codes: G06F 16/182 Distributed file systems
• CPC Class Codes: G06F 16/1827 Management specifically adapted to NAS management of storage area networks [SAN] G06F3/067
• CPC Class Codes: G06F 16/183 Provision of network file services by network file servers, e.g. by using NFS, CIFS network file access protocols H04L67/1097
• CPC Class Codes: G06F 2201/85 Active fault masking without idle spares active fault masking without idle spare hardware where processing functionality is redundant G06F11/2035

Abstract

The disclosed embodiments provide a system that adjusts the characteristics of a distributed filesystem using a locality policy. Two or more cloud controllers collectively manage distributed filesystem data that is stored in one or more cloud storage systems; the cloud controllers cache and ensure data consistency for the stored data. During operation, a cloud controller receives a locality policy that specifies one or more management policies for the cloud controller. The portion of the distributed filesystem'"'"'s data that is managed, created, and/or cached at the cloud controller is then managed based on this locality policy. Locality policies facilitate customizing and optimizing data management for the distributed filesystem to fit the needs of an organization (e.g., specific sets of users, applications, and/or datasets).

Claims

1. A computer-implemented method for customizing data management for a distributed filesystem, the method comprising:

collectively managing the data of the distributed filesystem using two or more cloud controllers, wherein collectively managing the data comprises;

upon receiving in a cloud controller new data from a client, sending from the cloud controller an incremental metadata snapshot for the new data, wherein the incremental metadata snapshot is received by the other cloud controllers of the distributed filesystem;

storing the data for the distributed filesystem in one or more cloud storage systems, wherein the cloud controllers cache and ensure data consistency for data stored in the cloud storage systems; and

sending an incremental data snapshot containing the new data from the cloud controller to a cloud storage system;

collectively presenting a unified namespace for the distributed filesystem to the clients of the distributed filesystem via the two or more cloud controllers, wherein the clients access the distributed filesystem via the cloud controllers, wherein the file data for the distributed filesystem is stored in the cloud storage system, wherein cloud controllers cache a subset of the file data from the remote cloud storage system that is being actively accessed by each respective cloud controller'"'"'s clients, wherein new file data received by each cloud controller from its clients is written to the remote cloud storage system, wherein the metadata hierarchy in the cloud controller tracks the location of distributed filesystem data blocks in the remote cloud storage system and cached distributed filesystem data blocks in the cloud controller, wherein the cloud controller uses the metadata hierarchy to locate and download requested, uncached data blocks in the distributed filesystem from the remote cloud storage system;

maintaining at each cloud controller a copy of the metadata for the files stored in the distributed filesystem, wherein changes to the metadata for the distributed filesystem are communicated to the set of cloud controllers for the distributed filesystem to ensure that the clients of the distributed filesystem share a consistent view of the files in the distributed filesystem;

receiving at the cloud controller a locality policy that specifies one or more management policies that are applied to all of a set of two or more cloud controllers, wherein the set of two or more cloud controllers includes the cloud controller;

wherein the locality policy specifies a high-priority file that should be pinned in a cache of the cloud controller;

and managing data for the distributed filesystem on the set of two or more at cloud controllers based on the locality policy;

wherein each cloud controller in the set of two or more cloud controllers downloads modified data for the high-priority file and pins the modified data for the high-priority file in the cloud controller'"'"'s local cache for the distributed filesystem, thereby ensuring that the modified data for the high-priority file is immediately available to all of the clients of the set of two or more cloud controllers; and

wherein the locality policy specifies one or more of a pre-condition that needs to be met before the cloud controller applies the locality policy and a time interval in which the cloud controller will apply the locality policy.

2. The computer-implemented method of claim 1,

wherein the scope of the locality policy is defined at the granularity of multiple cloud controllers of the distributed filesystem;

and

wherein the ability to specify locality policies that span a range of granularities facilitates customizing caching behavior for the distributed fileystem to meet the needs of one or more of an organization, a set of users, a specific application, and a specific dataset.

3. The computer-implemented method of claim 2, wherein the method further comprises:

determining in the cloud controller that a second incremental metadata snapshot distributed by a second cloud controller includes modified metadata for the high-priority file, wherein the modified metadata specifies a cloud file that stores modified data for the high-priority file;

preemptively downloading the cloud file to the cloud controller; and

pinning the modified data for the high-priority file in cache; and

wherein preemptively downloading and pinning data for high-priority files facilitates improving access performance for high-priority files.

4. The computer-implemented method of claim 2,

wherein the locality policy specifies that a set of files are associated with each other;

and

wherein the associated files share at least one of a file type and a common usage characteristic.

5. The computer-implemented method of claim 4, wherein the locality policy specifies a block size for the set of associated files that matches a preferred application block size to improve a deduplication hit rate for the set of files, wherein different applications may be associated with different preferred application block sizes.

6. The computer-implemented method of claim 4,

wherein the locality policy specifies where data and metadata will be stored in a local rotating disk drive in the cloud controller;

and

wherein the locality policy ensures that data and metadata that are expected to be accessed most frequently are stored in an outer edge portion of a disk platter of the rotating disk drive, thereby ensuring that the frequently-accessed data can be accessed with the highest performance.

7. The computer-implemented method of claim 4,

wherein the method further comprises storing the associated files in the same cloud file;

and

wherein writing associated files into the same cloud file facilitates subsequently performing a uniform operation on the associated files.

8. The computer-implemented method of claim 7, wherein the uniform operation comprises one or more of the following:

transferring the cloud file from a primary cloud storage system to a secondary cloud storage system;

storing the cloud file initially to the secondary cloud storage system instead of to the primary cloud storage system; and

transferring the cloud file from a first cloud storage system to an archival cloud storage system.

9. The computer-implemented method of claim 8, wherein the locality policy specifies a time interval after which the cloud controller will perform the uniform action operation.

10. The computer-implemented method of claim 9, wherein the locality policy specifies a pre-condition that needs to be met before the cloud controller performs the uniform operation.

11. The computer-implemented method of claim 2,

wherein the locality policy specifies a peer cloud controller for the cloud controller;

and

wherein the method further comprises;

receiving at the cloud controller a request from the client for a data block for a file in the distributed filesystem;

determining that the requested data block is not currently cached in the cloud controller; and

sending a peer cache request for the requested data block from the cloud controller to the peer cloud controller.

12. The computer-implemented method of claim 2,

wherein the locality policy specifies a level of aggressiveness for pre-fetching decisions in the cloud controller;

wherein aggressive pre-fetching decisions reduce average file access latency for a set of clients accessing files in the distributed filesystem via the cloud controller by speculatively pre-fetching additional cloud files that include data related to a current request; and

wherein less aggressive pre-fetching decisions reduce network bandwidth usage for the cloud controller by performing fewer speculative pre-fetch operations, thereby potentially increasing average file access latency for the set of clients.

13. The computer-implemented method of claim 2, wherein the locality policy specifies that a file received by the cloud controller from the client should remain limited to the scope of the cloud controller and not be added to the distributed filesystem, and hence should not be stored to the cloud storage system nor shared in the distributed filesystem.

14. The computer-implemented method of claim 2,

wherein the locality policy is received subsequent to storing a set of data and metadata in the distributed filesystem;

and

wherein applying the locality policy to data and metadata that were previously stored in the distributed filesystem comprises analyzing the metadata of the distributed filesystem, downloading a set of existing cloud files that contain data affected by the locality policy, creating new cloud files containing this previously stored data that are organized to reflect the locality policy, and uploading these new cloud files to the cloud storage system to replace the existing cloud files.

15. The computer-implemented method of claim 2,

wherein the locality policy specifies tracking all accesses for a file in the distributed filesystem;

and

using the tracking data for the file to audit the usage and tune the performance of the distributed filesystem.

16. A non-transitory computer-readable storage medium storing instructions that when executed by a computer cause the computer to perform a method for customizing data management for a distributed filesystem, the method comprising:

collectively managing the data of the distributed filesystem using two or more cloud controllers, wherein collectively managing the data comprises;

upon receiving in a cloud controller new data from a client, sending from the cloud controller an incremental metadata snapshot for the new data, wherein the incremental metadata snapshot is received by the other cloud controllers of the distributed filesystem;

storing the data for the distributed filesystem in one or more cloud storage systems, wherein the cloud controllers cache and ensure data consistency for data stored in the cloud storage systems; and

sending an incremental data snapshot containing the new data from the cloud controller to a cloud storage system;

collectively presenting a unified namespace for the distributed filesystem to the clients of the distributed filesystem via the two or more cloud controllers, wherein the clients access the distributed filesystem via the cloud controllers, wherein the file data for the distributed filesystem is stored in the cloud storage system, wherein cloud controllers cache a subset of the file data from the remote cloud storage system that is being actively accessed by each respective cloud controller'"'"'s clients, wherein new file data received by each cloud controller from its clients is written to the remote cloud storage system, wherein the metadata hierarchy in the cloud controller tracks the location of distributed filesystem data blocks in the remote cloud storage system and cached distributed filesystem data blocks in the cloud controller, wherein the cloud controller uses the metadata hierarchy to locate and download requested, uncached data blocks in the distributed filesystem from the remote cloud storage system;

maintaining at each cloud controller a copy of the metadata for the files stored in the distributed filesystem, wherein changes to the metadata for the distributed filesystem are communicated to the set of cloud controllers for the distributed filesystem to ensure that the clients of the distributed filesystem share a consistent view of the files in the distributed filesystem;

receiving at the cloud controller a locality policy that specifies one or more management policies far to be applied across a set of two or more cloud controllers, wherein the set of two or more cloud controllers includes the cloud controller; and

managing data for the distributed filesystem on the set of two or more a cloud controllers based on the locality policy;

wherein each cloud controller in the set of two or more cloud controllers downloads modified data for the high-priority file and pins the modified data for the high-priority file in the cloud controller'"'"'s local cache for the distributed filesystem, thereby ensuring that the modified data for the high-priority file is immediately available to all of the clients of the set of two or more cloud controllers; and

wherein the locality policy specifies one or more of a pre-condition that needs to be met before the cloud controller applies the locality policy and a time interval in which the cloud controller will apply the locality policy.

17. A cloud controller that performs customized data management for a distributed filesystem, comprising:

a processor;

a storage mechanism that stores metadata for the distributed filesystem; and

a storage management mechanism;

wherein two or more cloud controllers collectively manage the data of the distributed filesystem, wherein collectively managing the data comprises;

upon receiving in a cloud controller new data from a client, sending from the cloud controller an incremental metadata snapshot for the new data, wherein the incremental metadata snapshot is received by the other cloud controllers of the distributed filesystem;

storing the data for the distributed filesystem in one or more cloud storage systems, wherein the storage management mechanisms of the cloud controllers are configured to cache and ensure data consistency for data stored in the cloud storage systems;

sending an incremental data snapshot containing the new data from the cloud controller to a first cloud storage system;

collectively presenting a unified namespace for the distributed filesystem to the clients of the distributed filesystem via the two or more cloud controllers, wherein the clients access the distributed filesystem via the cloud controllers, wherein the file data for the distributed filesystem is stored in the cloud storage system, wherein cloud controllers cache a subset of the file data from the remote cloud storage system that is being actively accessed by each respective cloud controller'"'"'s clients, wherein new file data received by each cloud controller from its clients is written to the remote cloud storage system;

wherein the metadata hierarchy in the cloud controller tracks the location of distributed filesystem data blocks in the remote cloud storage system and cached distributed filesystem data blocks in the cloud controller, wherein the cloud controller uses the metadata hierarchy to locate and download requested, uncached data blocks in the distributed filesystem from the remote cloud storage system; and

maintaining at each cloud controller a copy of the metadata for the files stored in the distributed filesystem, wherein changes to the metadata for the distributed filesystem are communicated to the set of cloud controllers for the distributed filesystem to ensure that the clients of the distributed filesystem share a consistent view of the files in the distributed filesystem; and

wherein the storage management mechanism is further configured to;

receive a locality policy that specifies one or more management policies to be applied across a set of two or more cloud controllers, wherein the set of two or more cloud controllers includes the cloud controller; and

manage data for the distributed filesystem at the cloud controller based on the locality policy;

wherein each cloud controller in the set of two or more cloud controllers downloads modified data for the high-priority file and pins the modified data for the high-priority file in the cloud controller'"'"'s local cache for the distributed filesystem, thereby ensuring that the modified data for the high-priority file is immediately available to all of the clients of the set of two or more cloud controllers; and

wherein the locality policy specifies one or more of a pre-condition that needs to be met before the cloud controller applies the locality policy and a time interval in which the cloud controller will apply the locality policy.