Consistency checker for global de-duplication clustered file system

Patent metadata

Field	Value
Publication number	US-10049118-B2
Application number	US-201514727005-A
Country	US
Kind code	B2
Filing date	Jun 1, 2015
Priority date	Jun 1, 2015
Publication date	Aug 14, 2018
Grant date	Aug 14, 2018

Abstract

A cluster-wide consistency checker ensures that two file systems of a storage input/output (I/O) stack executing on each node of a cluster are self-consistent as well as consistent with respect to each other. The file systems include a deduplication file system and a host-facing file system that cooperate to provide a layered file system of the storage I/O stack. The deduplication file system is a log-structured file system managed by an extent store layer of the storage I/O stack, whereas the host-facing file system is managed by a volume layer of the stack. Illustratively, each log-structured file system implements a key-value store and cooperates with other nodes of the cluster to provide a cluster-wide (global) key-value store. The consistency checker verifies and/or fixes on-disk structures of the layered file system to ensure its consistency. To that end, the consistency checker may determine whether there are inconsistencies in the key-value store and, if so, reconciles those inconsistencies from a client (volume layer) perspective.

First claim

What is claimed is:

1. A method comprising:
checking self-consistency of first on-disk data structures of an extent store, in response to recovery of a layered file system, the extent store included as a first layer of the layered file system serviced by a storage system coupled to one or more storage devices storing extents of the extent store, the first on-disk data structures of the extent store including one or more hash tables configured to maintain an extent key and a reference count for each of the extents;
checking self-consistency of second on-disk data structures of a volume layer in response to the recovery of the layered file system, the volume layer included as a second layer of the layered file system, the second on-disk structures of the volume layer including one or more dense tree structures configured to maintain mappings from host-accessible logical unit number (LUN) addresses to the extent keys; and
reconciling, for each extent key in the volume layer mappings, the reference count of a same extent key maintained in the extent store.

2. The method of claim 1 wherein checking the self-consistency of the extent store further comprises repairing the reference count for a de-duplicated extent.

3. The method of claim 1 wherein checking the self-consistency of the extent store further comprises validating an addressable location for each extent of the extent store.

4. The method of claim 1 further comprising:
publishing each extent key maintained in the volume layer mappings to a first set of map reducer bins, wherein each map reducer bin is a key-value store maintaining the extent key as a key of the key-value store and the reference count as a value of the key-value store.

5. The method of claim 4 further comprising:
merging the first set of map reducer bins associated with a first host-accessible LUN with a second set of map reducer bins associated with a second host-accessible LUN; and
reconciling for each extent key of a merged result of the first and second sets of map reducer bins, the reference count of the same extent key maintained in the extent store.

6. The method of claim 4 wherein publishing the extent key maintained in the volume layer to the first set of map reducer bins increases the reference count of the same extent key maintained in the key-value store of the map reducer bin.

7. The method of claim 4 further comprising:
publishing to the first set of map reducer bins, a refcount log extent key maintained in a refcount log for unreferencing an extent of the extent store, wherein update to the extent store from the refcount log is deferred; and
increasing the reference count of the extent key for the unreferenced extent maintained in the key-value store of the map reducer bin.

8. The method of claim 1 wherein the checked extent store is written to the storage devices.

9. The method of claim 4 wherein the mappings of the volume layer are maintained in a plurality of regions, and wherein each extent key maintained in the volume layer mappings per region is published in parallel with other regions.

10. The method of claim 1 wherein reconciling each extent key in the volume layer mappings is performed while the layered file system is offline.

11. A system comprising:
a storage array coupled to a storage system, the storage system having a memory connected to a processor via a bus;
a storage input/output (I/O) stack executing on the processor of the storage system, the storage I/O to stack configured to:
check self-consistency of first on-disk data structures of an extent store, in response to recovery of a layered file system, the extent store included as a first layer of the layered file system executed by the storage I/O stack, the storage array storing extents of the extent store, the first on-disk data structures of the extent store configured to maintain an extent key and a reference count for each of the extents;
check self-consistency of second on-disk data structures of a volume layer in response to the recovery of the layered file system, the volume layer included as a second layer of the layered file system, the second on-disk structures of to the volume layer including one or more dense tree structures configured to maintain mappings from host-accessible logical unit number (LUN) addresses to the extent keys; and
reconcile, for each extent key in the volume layer mappings, the reference count of a same extent key maintained in the extent store.

12. The system of claim 11 wherein checking the self-consistency of the extent store further comprises repairing the reference count for a de-duplicated extent.

13. The system of claim 11 wherein checking the self-consistency of the extent store further comprises validating an addressable location for each extent of the extent store.

14. The system of claim 11 wherein the storage I/O stack is further configured to:
publish each extent key maintained in the volume layer mappings to a first set of map reducer bins, wherein each map reducer bin is a key-value store maintaining the extent key as a key of the key-value store and the reference count as a value of the key-value store.

15. The system of claim 14 wherein the storage I/O stack is further configured to:
merge the first set of map reducer bins associated with a first host-accessible LUN with a second set of map reducer bins associated with a second host-accessible LUN; and
reconcile, for each extent key of a merged result of the first and second sets of map reducer bins, the reference count of the same extent key maintained in the extent store.

16. The system of claim 14 wherein publishing the extent key maintained in the volume layer to the first set of map reducer bins increases the reference count of the same extent key maintained in the key-value store of the map reducer bin.

17. The system of claim 14 wherein the storage I/O stack is further configured to:
publish to the first set of map reducer bins, a refcount log extent key maintained in a refcount log for unreferencing an extent of the extent store, wherein update to the extent store from the refcount log is deferred; and
increase the reference count of an extent key for the unreferenced extent maintained in the key-value store of the map reducer bin.

18. The system of claim 11 wherein the checked extent store is written to the storage array.

19. The system of claim 14 wherein the mappings of the volume layer are maintained in a plurality of regions, and wherein each extent key maintained in the volume layer mappings per region is published in parallel with other regions.

20. A non-transitory computer readable storage medium containing executable program instructions for execution by a processor included in a storage system having a memory, the storage system coupled to one or more storage devices, comprising program instructions that:
check self-consistency of first on-disk data structures of an extent store, in response to recovery of a layered file system, the extent store included as a first layer of the layered file system, the one or more storage devices storing extents of the extent store, the first on-disk data structures of the extent store including one or more hash tables configured to maintain an extent key and a reference count for each of the extents;
check self-consistency of second on-disk data structures of a volume layer in response to the recovery of the layered file system, the volume layer included as a second layer of the layered file system, the second on-disk structu

Assignees

• Netapp Inc

Inventors

• Patel Dhaval
• Patel Chaitanya
• Muth John
• Krishnamachari Srinath

Classifications

• G06F16/1748 Primary

  De-duplication implemented within the file system, e.g. based on file segments (de-duplication techniques in storage systems for the management of data blocks G06F3/0641) · CPC title

• G06F17/30156 Primary

  Physics · mapped topic