Serverless solution for continuous data protection

1. A system comprising: a processor; and memory configured to store one or more sequences of instructions which, when executed by the processor, cause the processor to carry out the steps of:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site;

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site, and

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored; and

managing data growth at the replica site by:

periodically creating a new database table for new metadata associated with new IOs;

consolidating, in conjunction with the periodically creating a new database table, old IOs into a volume image; and

deleting an old database table storing old metadata associated with the old IOs, and old objects in which the old IOs are stored.

2. The system of claim 1 wherein the processor further carries out the steps of:

tracking a total size of the plurality of IOs that have been aggregated; and

performing the transmitting when the total size exceeds a predetermined threshold size.

3. The system of claim 1 wherein the processor further carries out the steps of:

transmitting a copy of the volume to the replica site.

4. The system of claim 1 wherein the processor further carries out the steps of:

maintaining, at the replica site, an initial copy of the volume;

receiving, at the replica site, a request to read a set of blocks as of a time T and having a starting offset X and a size Y;

querying the database table to obtain a list of IOs having write timestamps before the time T and writes falling within a boundary defined by the starting offset X and the size Y;

selecting, from the list of IOs, IOs that last wrote to blocks within the boundary;

using the selected IOs, cross-referencing corresponding object IDs listed in the database table, the object IDs identifying objects storing writes of the selected IOs;

reading the identified objects to return first data responsive to the request and stored in the identified objects, the first data comprising data that changed as of the time T; and

reading the initial snapshot copy of the volume to return second data responsive to the request and stored in the initial copy of the volume, the second data comprising data that remained unchanged as of the time T.

5. A system comprising: a processor; and memory configured to store one or more sequences of instructions which, when executed by the processor, cause the processor to carry out the steps of:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site;

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site,

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored, and

wherein the database table is a first database table and the processor further carries out the steps of:

dividing the volume into a plurality of regions;

associating a first region with the first database table;

associating a second region with a second database table;

intercepting a first IO written to the volume, the first IO starting in the first region and ending in the second region;

separating the first IO into an IO first portion and an IO second portion;

generating first metadata associated with the IO first portion; and

generating second metadata associated with the IO second portion, wherein the first metadata is stored in the first database table and comprises a first write timestamp indicating when the first IO was written, a first write offset corresponding to an original write offset of the first IO, and a first write size equal to a size of the first region minus the first write offset, and

wherein the second metadata is stored in the second database table and comprises the same first write timestamp, a second write offset starting at a beginning offset of the second region, and a second write size equal to a remainder of the first IO, the remainder being equal to an original write size of the first IO minus the first write size.

6. A method comprising:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site;

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site, and

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored; and

managing data growth at the replica site by:

periodically creating a new database table for new metadata associated with new IOs;

consolidating, in conjunction with the periodically creating a new database table, old IOs into a volume image; and

deleting an old database table storing old metadata associated with the old IOs, and old objects in which the old IOs are stored.

7. The method of claim 6 comprising:

tracking a total size of the plurality of IOs that have been aggregated; and

performing the transmitting when the total size exceeds a predetermined threshold size.

8. The method of claim 6 comprising:

transmitting a copy of the volume to the replica site.

9. The method of claim 6 comprising:

maintaining, at the replica site, an initial copy of the volume;

receiving, at the replica site, a request to read a set of blocks as of a time T and having a starting offset X and a size Y;

querying the database table to obtain a list of IOs having write timestamps before the time T and writes falling within a boundary defined by the starting offset X and the size Y;

selecting, from the list of IOs, IOs that last wrote to blocks within the boundary;

using the selected IOs, cross-referencing corresponding object IDs listed in the database table, the object IDs identifying objects storing writes of the selected IOs;

reading the identified objects to return first data responsive to the request and stored in the identified objects, the first data comprising data that changed as of the time T; and

reading the initial snapshot copy of the volume to return second data responsive to the request and stored in the initial copy of the volume, the second data comprising data that remained unchanged as of the time T.

10. A method comprising:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site; and

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site,

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored, and

wherein the database table is a first database table and the method comprises:

dividing the volume into a plurality of regions;

associating a first region with the first database table;

associating a second region with a second database table;

intercepting a first IO written to the volume, the first IO starting in the first region and ending in the second region;

separating the first IO into an IO first portion and an IO second portion;

generating first metadata associated with the IO first portion; and

generating second metadata associated with the IO second portion, wherein the first metadata is stored in the first database table and comprises a first write timestamp indicating when the first IO was written, a first write offset corresponding to an original write offset of the first IO, and a first write size equal to a size of the first region minus the first write offset, and

wherein the second metadata is stored in the second database table and comprises the same first write timestamp, a second write offset starting at a beginning offset of the second region, and a second write size equal to a remainder of the first IO, the remainder being equal to an original write size of the first IO minus the first write size.

11. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors to implement a method comprising:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site;

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site, and

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored; and

managing data growth at the replica site by:

periodically creating a new database table for new metadata associated with new IOs;

consolidating, in conjunction with the periodically creating a new database table, old IOs into a volume image; and

deleting an old database table storing old metadata associated with the old IOs, and old objects in which the old IOs are stored.

12. The computer program product of claim 11 wherein the method comprises:

tracking a total size of the plurality of IOs that have been aggregated; and

performing the transmitting when the total size exceeds a predetermined threshold size.

13. The computer program product of claim 11 wherein the method comprises:

transmitting a copy of the volume to the replica site.

14. The computer program product of claim 11 wherein the method comprises:

maintaining, at the replica site, an initial copy of the volume;

receiving, at the replica site, a request to read a set of blocks as of a time T and having a starting offset X and a size Y;

querying the database table to obtain a list of IOs having write timestamps before the time T and writes falling within a boundary defined by the starting offset X and the size Y;

selecting, from the list of IOs, IOs that last wrote to blocks within the boundary;

using the selected IOs, cross-referencing corresponding object IDs listed in the database table, the object IDs identifying objects storing writes of the selected IOs;

reading the identified objects to return first data responsive to the request and stored in the identified objects, the first data comprising data that changed as of the time T; and

reading the initial snapshot copy of the volume to return second data responsive to the request and stored in the initial copy of the volume, the second data comprising data that remained unchanged as of the time T.

15. A computer program product, comprising a non-transitory computer-readable medium having a computer-readable program code embodied therein, the computer-readable program code adapted to be executed by one or more processors to implement a method comprising:

intercepting a plurality of input/output (IOs) written to a volume of a block storage device at a production site;

aggregating, at the production site, the plurality of IOs for storage in an object at a replica site, remote from the production site;

transmitting, from the production site to the replica site, the aggregated plurality of IOs and metadata associated with each IO, the metadata for each IO comprising a write timestamp indicating when an IO was written, a write offset of the IO, and a write size of the IO,

wherein the aggregated plurality of IOs are stored in a single object maintained in an object storage at the replica site,

wherein upon the single object being stored in the object storage, a function is triggered that writes entries to a database table maintained at the replica site, each entry corresponding to a respective IO of the aggregated plurality of IOs, and comprising a write timestamp, write offset, and write size of the respective IO, and an object identifier (ID) of the single object in which the aggregated plurality of IOs are stored, and

wherein the database table is a first database table and the method further comprises:

dividing the volume into a plurality of regions;

associating a first region with the first database table;

associating a second region with a second database table;

intercepting a first IO written to the volume, the first IO starting in the first region and ending in the second region;

separating the first IO into an IO first portion and an IO second portion;

generating first metadata associated with the IO first portion; and

generating second metadata associated with the IO second portion, wherein the first metadata is stored in the first database table and comprises a first write timestamp indicating when the first IO was written, a first write offset corresponding to an original write offset of the first IO, and a first write size equal to a size of the first region minus the first write offset, and

wherein the second metadata is stored in the second database table and comprises the same first write timestamp, a second write offset starting at a beginning offset of the second region, and a second write size equal to a remainder of the first IO, the remainder being equal to an original write size of the first IO minus the first write size.