Error recovery in a storage cluster

• US 9,495,255 B2
• Filed: 08/07/2014
• Issued: 11/15/2016
• Est. Priority Date: 08/07/2014
• Status: Expired due to Fees

• Current Assignee: Pure Storage, Inc.
• Original Assignee: Pure Storage, Inc.
• Inventors: Davis, John D., Hayes, John, Tan, Zhangxi, Kannan, Hari, Miladinovic, Nenad
• Primary Examiner(s): KO, CHAE M
• Application Number:
  US14/454,501
• Publication Number:
  US 20160041887A1
• Time in Patent Office: 831 Days
• Field of Search: 714/6.2, 714/6.24
• US Class Current: 1/1
• CPC Class Codes: G06F 11/1068 in sector programmable memories, e.g. flash disk G06F11/1072 takes precedence
• CPC Class Codes: G06F 11/1072 in multilevel memories
• CPC Class Codes: G06F 11/1076 Parity data used in redundant arrays of independent storages, e.g. in RAID systems
• CPC Class Codes: G06F 11/1612 where the redundant component is persistent storage
• CPC Class Codes: G06F 11/2069 Management of state, configuration or failover
• CPC Class Codes: G06F 2201/805 Real-time
• CPC Class Codes: G06F 2201/82 Solving problems relating to consistency ensuring consistency in mirrored systems G06F11/2064
• CPC Class Codes: G06F 3/0619 in relation to data integrity, e.g. data losses, bit errors
• CPC Class Codes: G06F 3/064 Management of blocks
• CPC Class Codes: G06F 3/067 Distributed or networked storage systems, e.g. storage area networks [SAN], network attached storage [NAS]
• CPC Class Codes: G06F 3/0685 Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays
• CPC Class Codes: G06F 3/0689 Disk arrays, e.g. RAID, JBOD

Abstract

A plurality of storage nodes within a single chassis is provided. The plurality of storage nodes is configured to communicate together as a storage cluster. The plurality of storage nodes has a non-volatile solid-state storage for user data storage. The plurality of storage nodes is configured to distribute the user data and metadata associated with the user data throughout the plurality of storage nodes, with erasure coding of the user data. The plurality of storage nodes is configured to recover from failure of two of the plurality of storage nodes by applying the erasure coding to the user data from a remainder of the plurality of storage nodes. The plurality of storage nodes is configured to detect an error and engage in an error recovery via one of a processor of one of the plurality of storage nodes, a processor of the non-volatile solid state storage, or the flash memory.

Claims

1. A plurality of storage nodes within a single chassis, comprising:

the plurality of storage nodes configurable to communicate together as a storage cluster, each of the plurality of storage nodes having a non-volatile solid-state storage for user data storage, the non-volatile solid state storage including flash memory, the plurality of storage nodes configured to distribute the user data and metadata associated with the user data throughout the plurality of storage nodes, with erasure coding of the user data;

the plurality of storage nodes configurable to recover from failure of two of the plurality of storage nodes by applying the erasure coding to reading the user data from a remainder of the plurality of storage nodes; and

the plurality of storage nodes configurable to detect an error and engage in an error recovery via one of a processor of one of the plurality of storage nodes, a processor of the non-volatile solid state storage, or the flash memory, wherein the plurality of storage nodes locates and accesses a mirrored remote procedure call cache in one of the plurality of storage nodes, responsive to a failure of a differing one of the plurality of storage nodes having a remote procedure call cache as mirrored by the mirrored remote procedure call cache.

2. The plurality of storage nodes of claim 1, wherein the flash memory applies error correction to data stored therein.

3. The plurality of storage nodes of claim 1, wherein the non-volatile solid-state storage retries a read of the flash memory.

4. The plurality of storage nodes of claim 1, wherein the storage node applies the erasure coding to data shards read from the plurality of storage nodes.

5. The plurality of storage nodes of claim 1, wherein the plurality of storage nodes rebuilds the user data onto the remainder of the plurality of storage nodes with a second erasure coding differing from the erasure coding of the plurality of storage nodes.

6. The plurality of storage nodes of claim 1, wherein the plurality of storage nodes includes reassigning ownership of a data segment to one of the plurality of storage nodes, responsive to a failure of a differing one of the plurality of storage nodes as an owner of the data segment.

7. The plurality of storage nodes of claim 1, wherein the plurality of storage nodes locate and access a redundant copy of a portion of the metadata in one of the plurality of storage nodes, responsive to a failure of a differing one of the plurality of storage nodes having a memory holding the portion of the metadata.

8. A storage cluster, comprising:

a plurality of storage nodes;

each of the plurality of storage nodes having a non-volatile solid-state storage with flash memory for storage of user data, the plurality of storage nodes configurable to distribute the user data, using erasure coding, throughout the plurality of storage nodes, the plurality of storage nodes configurable to distribute metadata associated with the user data, and at least one redundant copy of the metadata, throughout the plurality of storage nodes, the plurality of storage nodes is configurable to recover from loss of a remote procedure cache by reading a mirrored remote procedure call cache; and

the plurality of storage nodes configurable to recover from error selected from a group consisting of;

reading user data;

reading the metadata;

loss of one or more storage nodes;

unavailability of a portion of the user data; and

unavailability of a portion of the meta data.

9. The storage cluster of claim 8, further comprising:

the plurality of storage nodes is configured to determine whether one of the plurality of storage nodes is unreachable and whether to rebuild user data responsive to a determination that the one of the plurality of storage nodes is unreachable.

10. The storage cluster of claim 8, further comprising:

the non-volatile solid-state storage is configured to recover from error in reading the flash memory by applying at least one retry to reading the flash memory.

11. The storage cluster of claim 8, further comprising:

the non-volatile solid-state storage is configured to recover from an error in reading the flash memory by directing the flash memory to apply one of a plurality of reference voltage levels or reference current levels to successive read retries of the flash memory.

12. The storage cluster of claim 8, further comprising:

the non-volatile solid-state storage is configured to recover from an error in reading the flash memory by applying a probabilistic calculation to multiple read retries of the flash memory.

13. A method for error recovery in a plurality of storage nodes having non-volatile solid-state storage, comprising:

distributing user data and metadata throughout the plurality of storage nodes through erasure coding, wherein the plurality of storage nodes are housed within a single chassis that couples the storage nodes as a cluster;

detecting an error within user data retrieved from one of the plurality of storage nodes;

determining one of;

recovering from the error by accessing the user data, via the erasure coding, from a remainder of the plurality of storage nodes, responsive to a determination that two of the plurality of storage nodes are unreachable;

recovering from the error by rebuilding the user data, via the erasure coding, into the remainder of the plurality of storage nodes, responsive to the determination that the two of the plurality of storage nodes are unreachable;

recovering from the error by retrying a read of flash memory of the non-volatile solid-state storage of one of the plurality of storage nodes, responsive to a determination that the error is a bit error, wherein retrying the read of the flash memory further comprises;

applying a probabilistic calculation as to whether a data bit is more likely to be a logical “

1”

value or a logical “

0”

value;

or

recovering from the error by reassigning ownership of a portion of the user data to a one of the plurality of storage nodes, responsive to a determination that a further one of the plurality of storage nodes, having ownership of the portion of the user data, is unreachable.

14. The method of claim 13, further comprising:

recovering from the error by reading a redundant copy of metadata in one of the plurality of storage nodes, responsive to a determination of loss of the metadata.

15. The method of claim 13, further comprising:

recovering from the error by reading a mirrored remote procedure call cache in one of the plurality of storage nodes, responsive to a determination of loss of a remote procedure call cache.

16. The method of claim 13, wherein retrying the read of the flash memory further comprises:

applying a plurality of reference voltages or reference currents in the flash memory.

17. The method of claim 13, further comprising:

recovering from the error by applying the erasure coding to read a data stripe across the plurality of storage nodes, responsive to a determination that the data stripe includes the error.