Preserving quality of service when replicating data objects

Patent metadata

Field	Value
Publication number	US-11023488-B1
Application number	US-201414576939-A
Country	US
Kind code	B1
Filing date	Dec 19, 2014
Priority date	Dec 19, 2014
Publication date	Jun 1, 2021
Grant date	Jun 1, 2021

Abstract

A technique for replicating a data object from a source data storage system (source) to a destination data storage system (destination) preserves QoS (Quality of Service) by transmitting QoS information from the source to the destination and applying the QoS information when storing data at the destination that arrive in replication instructions sent from the source. The QoS information indicates desired storage tiering for the data object at the source. In the event of a failure at the source, which results in failover from the source to the destination, the destination is able to provide users with the same quality of service as was desired at the source.

First claim

What is claimed is:

1. A method of preserving quality of service (QoS) when replicating data objects from a source data storage system (source) to a destination data storage system (destination), the method comprising:
generating, by the source, storage tiering information indicating desired storage tiering for a data object made available by the source to a set of hosts, the desired storage tiering indicating desired performance levels of storage drives on which the data object is placed;
transmitting the storage tiering information from the source to the destination, the destination thereby receiving the storage tiering information; and
in response to receiving replication instructions from the source that specify data to be written to a replica of the data object in the destination to promote keeping the replica current with the data object in the source, storing the data specified in the replication instructions in the replica on a set of media that meet the desired storage tiering indicated for the data object in the storage tiering information, such that storage tiering for the replica at the destination conforms with the desired storage tiering for the data object at the source,
wherein generating the storage tiering information includes generating a QoS map that associates each of a set of ranges of the data object with a respective storage tier, and wherein transmitting the storage tiering information from the source to the destination includes sending the QoS map from the source to the destination.

2. The method of claim 1 , wherein sending the QoS map from the source to the destination is performed prior to the destination receiving the replication instructions.

3. The method of claim 1 , wherein each of the replication instructions is part of a respective snapshot shipping operation between the source and the destination, the snapshot shipping operation including:
generating, by the source, a first snap of the data object at a first time and a second snap of the data object at a second time;
identifying a set of blocks that are different between the first snap and the second snap;
sending the set of blocks from the source to the destination, and
applying, by the destination, the set of blocks to the replica to make the replica content-consistent with the data object as of the time that the second snap was generated,
wherein each of the set of blocks is located within one of the set of ranges that the QoS map associates with a respective storage tier, and wherein applying the set of blocks to the replica includes (i) the destination accessing the QoS map and (ii) the destination storing each of the set of blocks on the storage tier that the QoS map associates with the range within which the respective block is located.

4. The method of claim 1 ,
wherein the replication instructions include continuous replication instructions for mirroring IO requests received by the source to the destination, the IO requests specifying data to be written to the data object on the source,
wherein storing the data specified in the replication instructions includes, for each mirrored IO request, (i) appending a tag to the mirrored IO request, the tag indicating a desired storage tiering for the data specified in the IO request as provided in the QoS map, (ii) providing the mirrored IO request, including the tag, to an IO stack on the destination, and (iii) storing, by the IO stack on the destination, the data specified by the mirrored IO request on the tier of storage indicated by the tag.

5. The method of claim 1 , wherein generating the storage tiering information includes specifying a single QoS for the data object as a whole, and wherein the method further comprises, for a second data object made available by the source:
generating, by the source, second storage tiering information indicating desired storage tiering for the second data object, the second storage tiering information specifying a single QoS for the second data object as a whole;
transmitting the second storage tiering information from the source to the destination, the destination thereby receiving the second storage tiering information; and
in response to receiving second replication instructions that specify data to be written to a replica of the second data object at the destination, storing the data specified in the second replication instructions in the replica on a set of media that meet the desired storage tiering indicated for the second data object in the second storage tiering information.

6. The method of claim 1 ,
wherein generating the storage tiering information for the data object includes (i) specifying a first storage tier for storing a first portion of the data object and (ii) specifying a second storage tier for storing a second portion of the data object,
wherein the data specified by the replication instructions include (i) a first set of data indicating changes in the first portion of the data object and (ii) a second set of data indicating changes in the second portion of the data object, and
wherein storing the data specified by the replication instructions in the replica includes (i) accessing the received storage tiering information, (ii) storing the first set of data on a first set of storage media providing first storage tier, and (iii) storing the second set of data on a second set of storage media providing the second storage tier.

7. The method of claim 6 ,
wherein the first set of data includes metadata of the data object and the first set of storage media include a set of flash drives, and
wherein the second set of data includes non-metadata data of the data object and the second set of storage media include a set of magnetic disk drives.

8. The method of claim 1 , wherein the replication instructions include instructions for performing multiple continuous replication operations, wherein receiving the storage tiering information from the source includes receiving a respective portion of the storage tiering information with each of the replication instructions, and wherein the method further comprises:
receiving, by the source, an IO request from one of the set of hosts, the IO request specifying data to be written to the data object; and
appending, by the source, a tag to the IO request, the tag specifying a desired QoS for storing the data specified by the IO request,
wherein transmitting the storage tiering information from the source to the destination includes the source mirroring the IO request, including the tag, to the destination, and
wherein storing the data specified in the replication instructions includes reading the tag by the destination and storing, by the destination, the data specified by the IO request in a set of storage media that meet the QoS specified in the tag.

9. The method of claim 1 , further comprising:
detecting, by the source, that a portion of the data object for which the storage tiering information specifies a slower tier of storage has become frequently accessed by the set of hosts;
in response to detecting the frequent access, changing the storage tiering information for the portion of the data object to specify a faster tier of storage;
transmitting the changed storage tiering information from the source to the destination; and
in response to receiving subsequent replication instructions from the source that specify data from the portion of the data object to be written to the replica, storing the specified data on a set of storage media that meet the changed storage tiering information specifying the faster tier of storage.

10. The method of claim 1 , wherein the source fails to provide all of the desired storage tiering for the data object specified in the st

Assignees

• Emc Corp
• Emc Ip Holding Co Llc

Inventors

• Bono Jean-Pierre
• Tummala Himabindu
• Davenport William C
• Natanzon Assaf

Classifications

• H04L67/1095 Primary

  Replication or mirroring of data, e.g. scheduling or transport for data synchronisation between network nodes · CPC title

• H04L67/61

  taking into account QoS or priority requirements · CPC title

• G06F16/1824

  implemented using Network-attached Storage [NAS] architecture (distributed or networked storage systems G06F3/067; protocols for distributed storage of data in a network H04L67/1097) · CPC title

• G06F3/0685

  Hybrid storage combining heterogeneous device types, e.g. hierarchical storage, hybrid arrays · CPC title

• G06F3/0647

  Migration mechanisms · CPC title