Write signature command

• US 8,898,112 B1
• Filed: 09/07/2011
• Issued: 11/25/2014
• Est. Priority Date: 09/07/2011
• Status: Active Grant

• Current Assignee: Emc IP Holding Company LLC (Dell Technologies Inc.)
• Original Assignee: EMC Corporation (Dell Technologies Inc.)
• Inventors: Natanzon, Assaf, Duprey, Dennis T., Owen, Karl M., Baruch, Leehod
• Primary Examiner(s): Uddin, Md. I
• Application Number:
  US13/226,546
• Time in Patent Office: 1,175 Days
• Field of Search: 707/634, 707/639, 707/640, 707/641, 707/656, 707/646
• US Class Current: 707/634
• CPC Class Codes: G06F 11/2074 Asynchronous techniques
• CPC Class Codes: G06F 11/2076 Synchronous techniques
• CPC Class Codes: G06F 11/2082 Data synchronisation
• CPC Class Codes: G06F 2201/83 the solution involving signatures
• CPC Class Codes: G06F 2201/855 Details of asynchronous mirroring using a journal to transfer not-yet-mirrored changes

Abstract

In one aspect, a method includes generating a write signature command. The write signature command is configured to write a signature to an offset in a storage array without data and to enable the storage array to write the data with the same signature to a volume if the data is available at the storage array.

Claims

1. A method, comprising:

generating a write signature command, the write signature command having parameters used in execution comprising an offset, a number of blocks of data and a signature, the write signature command configured to;

write the signature to the offset in a storage array without the data, the signature being a hash value generated using a hash function on the data; and

enable the storage array to write the data with the same signature to a volume if the data is available at the storage array;

executing the write signature command to synchronize data between a second site and a first site, executing the write signature command comprising;

reading a first hash value having a first offset at the first site associated with a block of data at the first site;

sending the first hash value to the second site;

using the write signature command with the first hash value received from the first site to determine if there is a block of data stored at the second site associated with the first hash value, wherein the write signature command uses the first hash value to copy a block of data, from a location at the second site indicated in a hash table associated with the first hash value, to a second offset at the second site if the first hash value is found in the hash table;

determining if the write signature command was successful by determining there is the block of data stored at the second site associated with the first hash value; and

if the write command is not successful;

reading the block of data from the first site associated with the first hash value; and

writing the block of data from the first site associated with the first hash value to the second offset at the second site, the second offset being equal to the first offset.

2. The method of claim 1, further comprising synchronizing data between a first site and a second site without sending the data from the first site to the second site if the second site already contains data from first site located in any location in the storage array.

3. The method of claim 1, further comprising:

receiving a write at the storage array;

reading a signature for the write;

invalidating a previous hash table entry for the signature;

determining a hash value for the write; and

updating the hash table.

4. The method of claim 3, further comprising:

receiving a write signature command containing a hash value,

searching for hash value in the hash table;

copying the data from the location indicated by the hash table to a location indicated by the write signature command if the hash value is found in the hash table; and

returning a mismatch message if the data is not found in the hash table.

5. The method of claim 1, further comprising:

intercepting a write command;

marking a write in a bitmap as dirty;

reading the data from offset indicated as dirty in the bit map;

calculating a signature for the data; and

adding the signature, for a hash table indicating the signature points to a write offset.

6. The method of claim 5, further comprising:

intercepting a write signature command;

searching for the hash value in the hash table;

copying the data from the location indicated by the hash table to a location indicated by the write signature command if the hash value is found and the hash value of the data indicated by the hash table is identical to the hash value; and

returning a mismatch message if the data is not found or is not identical.

7. The method of claim 1, further comprising using the write signature command to save bandwidth when transferring deduplicated data over a network.

8. The method of claim 7 wherein using the write signature command to save bandwidth when transferring deduplicated data over a network comprises:

generating, at the first site, the first signature associated with data to be sent to the second site;

sending the first signature to the second site;

attempting to write the first signature to a location at a storage LUN at the second site;

reading the data from the location at the storage LUN at the second site if the attempt to write the first signature was successful;

sending a status message indicating success if the attempting to write the first signature was successful;

sending a status message indicating failure if the attempting to write the first signature was unsuccessful;

sending the data from the first site to the second site and writing the data to an offset at the second site if the attempting to write the first signature was unsuccessful.

9. The method of claim 1, further comprising saving bandwidth in a snapshot shipping replication environment.

10. The method of claim 9 wherein saving bandwidth in a snapshot shipping replication environment comprises:

generating a first snapshot at a first site;

tracking changes occurring at the first site;

performing an initial sweep from the first snapshot to the second site;

erasing the first snapshot;

generating a second snapshot at the first site;

transferring the changes from the first site to the second site; and

renaming the second snapshot to be the first snapshot,

where transferring the changes comprises;

reading signatures of the changed locations;

writing signatures to a volume at the replica site using the signatures read at the first site; and

on failure reading the data from the location at the first site and writing the data to the changed locations at the replica site.

11. An article comprising:

a non-transitory machine-readable medium that stores executable instructions, the instructions causing a machine to;

generate a write signature command, the write signature command having parameters used in execution comprising an offset, a number of blocks of data and a signature, the write signature command configured to;

write the signature to the offset in a storage array without the data, the signature being a hash value generated using a hash function on the data; and

enable the storage array to write the data with the same signature to a volume if the data is available at the storage array;

execute the write signature command to synchronize data between a second site and a first site, the instructions causing a machine to execute the write signature command comprising instructions causing a machine to;

read a first hash value having a first offset at the first site associated with a block of data at the first site;

send the first hash value to the second site;

use the write signature command with the first hash value received from the first site to determine if there is a block of data stored at the second site associated with the first hash value, wherein the write signature command uses the first hash value to copy a block of data, from a location at the second site indicated in a hash table associated with the first hash value, to a second offset at the second site if the first hash value is found in the hash table;

determine if the write signature command was successful by determining there is the block of data stored at the second site associated with the first hash value; and

if the write command is not successful;

read the block of data from the first site associated with the first hash value; and

write the block of data from the first site associated with the first hash value to the second offset at the second site, the second offset being equal to the first offset.

12. The article of claim 11, further comprising instructions causing the machine to synchronize data between a first site and a second site without sending the data from the first site to the second site if the second site already contains data from first site located in any location in the storage array.

13. The article of claim 11, further comprising instructions causing the machine to:

receive a write at the storage array;

read a signature for the write;

invalidate a previous hash table entry for the signature;

determine a hash value for the write; and

update the hash table.

14. An apparatus, comprising:

hardware circuitry to generate a write signature command, the write signature command configured to;

generate a write signature command, the write signature command having parameters used in execution comprising an offset, a number of blocks of data and a signature, the write signature command configured to;

write the signature to the offset in a storage array without the data, the signature being a hash value generated using a hash function on the data; and

enable the storage array to write the data with the same signature to a volume if the data is available at the storage array;

execute the write signature command to synchronize data between a second site and a first site, the circuitry configured to execute the write signature command comprising circuitry configured to;

read a first hash value having a first offset at the first site associated with a block of data at the first site;

send the first hash value to the second site;

use the write signature command with the first hash value received from the first site to determine if there is a block of data stored at the second site associated with the first hash value, wherein the write signature command uses the first hash value to copy a block of data, from a location at the second site indicated in a hash table associated with the first hash value, to a second offset at the second site if the first hash value is found in the hash table;

determine if the write signature command was successful by determining there is the block of data stored at the second site associated with the first hash value; and

if the write command is not successful;

read the block of data from the first site associated with the first hash value; and

write the block of data from the first site associated with the first hash value to the second offset at the second site, the second offset being equal to the first offset.

15. The apparatus of claim 14 wherein the circuitry comprises at least one of a processor, a memory, programmable logic and logic gates.

16. The apparatus of claim 14, further comprising circuitry to synchronize data between a first site and a second site without sending the data from the first site to the second site if the second site already contains data from first site located in any location in the storage array.

17. The apparatus of claim 14, further comprising circuitry to:

receive a write at the storage array;

read a signature for the write;

invalidate a previous hash table entry for the signature;

determine a hash value for the write; and

update the hash table.