Smart data offload sync replication
A method is provided, comprising: receiving, from a host device, a request to create a token that represents one or more data items that are stored in a first volume, the first volume being stored in a source system; estimating an I/O latency limit for the first volume; estimating a token creation time for the token; comparing the token creation time to the I/O latency limit; when the token creation time matches the I/O latency limit: creating the token, creating a snapshot pair, mapping the token to the snapshot pair, and returning the token to the host device; when the token creation time does not match the I/O latency limit: creating the token, creating a first snapshot of the first volume at the source system, mapping the first snapshot to the token, and returning the token to the host device.
1. A method, comprising:
receiving, from a host device, a request to create a token that represents one or more data items that are stored in a first volume, the first volume being stored in a source system;
estimating an input-output (I/O) latency limit for the first volume;
estimating a token creation time for the token, the token creation time being an amount of time that is needed to create the token;
comparing the token creation time to the I/O latency limit;
when the token creation time matches the I/O latency limit: creating the token, creating a snapshot pair, mapping the token to the snapshot pair, and returning the token to the host device, the snapshot pair including a first snapshot of the first volume that is created at the source system and a second snapshot of a replica of the first volume that is stored at a target system; and
when the token creation time does not match the I/O latency limit: creating the token, creating a first snapshot of the first volume at the source system, mapping the first snapshot to the token, and returning the token to the host device.
2. The method of claim 1 , wherein no snapshots of the replica of the first volume are created in response to the request to create a token, when the token creation time does not match the I/O latency limit.
3. The method of claim 1 , wherein the I/O latency limit is estimated based on at least one of: (i) a QoS configuration of the source system, (ii) an application profile that is associated with the first volume, or (iii) a workload on the source system.
4. The method of claim 1 , wherein the token creation time is estimated based on one or more of: (i) time to suspend and drain pending I/Os that are associated with the first volume, (ii) historical data associated with one or more snapshots of the first volume that were created in the past, and (iii) historical data associated with one or more snapshots of the replica of the first volume that were created in the past.
5. The method of claim 1 , wherein:
creating the snapshot pair includes initiating a process for creating the snapshot pair,
the snapshot pair is created only when no conditions for terminating the process are detected during execution of the process, and
when a condition for terminating the process is detected during the execution of the process, the process is aborted, and only the first snapshot of the first volume is created and mapped to the token.
6. The method of claim 1 , further comprising:
receiving, at the source system, a first write-via-token command that is associated with the token, the first write-via-token command instructing the source system to copy the one or more data items from the first volume to a second volume;
detecting whether the token is mapped to a snapshot pair or to a single snapshot;
when the token is mapped to a snapshot pair: copying the one or more data items from the first volume to the second volume, and transmitting to the target system a second write-via-token command that instructs the target system to copy to the one or more data items from the replica of the first volume to a replica of the second volume; and
when the token is mapped to a single snapshot: copying the one or more data items from the first volume to the second volume; and transmitting to the target system one or more write-via-hash commands that instruct the target system to copy to the one or more data items from the replica of the first volume to the replica of the second volume.
7. The method of claim 1 , wherein the source system and the target system are part of a content-addressable storage system.
8. A computing system, comprising:
a memory; and
at least one processor that is operatively coupled to the memory, the at least one processor being configured to perform the operations of:
receiving, from a host device, a request to create a token that represents one or more data items that are stored in a first volume, the first volume being stored in a source system;
estimating an input-output (I/O) latency limit for the first volume;
estimating a token creation time for the token, the token creation time being an amount of time that is needed to create the token;
comparing the token creation time to the I/O latency limit;
when the token creation time matches the I/O latency limit: creating the token, creating a snapshot pair, mapping the token to the snapshot pair, and returning the token to the host device, the snapshot pair including a first snapshot of the first volume that is created at the source system and a second snapshot of a replica of the first volume that is stored at a target system; and
when the token creation time does not match the I/O latency limit: creating the token, creating a first snapshot of the first volume at the source system, mapping the first snapshot to the token, and returning the token to the host device.
9. The computing system of claim 8 , wherein no snapshots of the replica of the first volume are created in response to the request to create a token, when the token creation time does not match the I/O latency limit.
10. The computing system of claim 8 , wherein the I/O latency limit is estimated based on at least one of: (i) a QoS configuration of the source system, (ii) an application profile that is associated with the first volume, or (iii) a workload on the source system.
11. The computing system of claim 8 , wherein the token creation time is estimated based on one or more of: (i) time to suspend and drain pending I/Os that are associated with the first volume, (ii) historical data associated with one or more snapshots of the first volume that were created in the past, and (iii) historical data associated with one or more snapshots of the replica of the first volume that were created in the past.
12. The computing system of claim 8 , wherein:
creating the snapshot pair includes initiating a process for creating the snapshot pair,
the snapshot pair is created only when no conditions for terminating the process are detected during execution of the process, and
when a condition for terminating the process is detected during the execution of the process, the process is aborted, and only the first snapshot of the first volume is created and mapped to the token.
13. The computing system of claim 8 , wherein the at least one processor is further configured to perform the operations of:
receiving, at the source system, a first write-via-token command that is associated with the token, the first write-via-token command instructing the source system to copy the one or more data items from the first volume to a second volume;
detecting whether the token is mapped to a snapshot pair or to a single snapshot;
when the token is mapped to a snapshot pair: copying the one or more data items from the first volume to the second volume, and transmitting to the target system a second write-via-token command that instructs the target system to copy to the one or more data items from the replica of the first volume to a replica of the second volume; and
when the token is mapped to a single snapshot: copying the one or more data items from the first volume to the second volume; and transmitting to the target system one or more write-via-hash commands that instruct the target system to copy to the one or more data items from the replica of the first volume to the replica of the second volume.
14. The computing system of claim 8 , wherein the source system and the target system are part of a content-addressable storage system, and the computing system is part of the source system.
15. A non-transitory computer-readable medium storing one or more processor-executable instructions, which, when executed by at least one processor, cause the at least one processor to perform the operations of:
receiving, from a host device, a request to create a token that represents one or more data items that are stored in a first volume, the first volume being stored in a source system;
estimating an input-output (I/O) latency limit for the first volume;
estimating a token creation time for the token, the token creation time being an amount of time that is needed to create the token;
comparing the token creation time to the I/O latency limit;
when the token creation time matches the I/O latency limit: creating the token, creating a snapshot pair, mapping the token to the snapshot pair, and returning the token to the host device, the snapshot pair including a first snapshot of the first volume that is created at the source system and a second snapshot of a replica of the first volume that is stored at a target system; and
when the token creation time does not match the I/O latency limit: creating the token, creating a first snapshot of the first volume at the source system, mapping the first snapshot to the token, and returning the token to the host device.
16. The non-transitory computer-readable medium of claim 15 , wherein no snapshots of the replica of the first volume are created in response to the request to create a token, when the token creation time does not match the I/O latency limit.
17. The non-transitory computer-readable medium of claim 15 , wherein the I/O latency limit is estimated based on at least one of: (i) a QoS configuration of the source system, (ii) an application profile that is associated with the first volume, or (iii) a workload on the source system.
18. The non-transitory computer-readable medium of claim 15 , wherein the token creation time is estimated based on one or more of: (i) time to suspend and drain pending I/Os that are associated with the first volume, (ii) historical data associated with one or more snapshots of the first volume that were created in the past, and (iii) historical data associated with one or more snapshots of the replica of the first volume that were created in the past.
19. The non-transitory computer-readable medium of claim 15 , wherein:
creating the snapshot pair includes initiating a process for creating the snapshot pair,
the snapshot pair is created only when no conditions for terminating the process are detected during execution of the process, and
when a condition for terminating the process is detected during the execution of the process, the process is aborted, and only the first snapshot of the first volume is created and mapped to the token.
20. The non-transitory computer-readable medium of claim 15 , wherein the one or more processor-executable instructions, when executed by the at least one processor, further cause the at least one processor to perform the operation of:
receiving, at the source system, a first write-via-token command that is associated with the token, the first write-via-token command instructing the source system to copy the one or more data items from the first volume to a second volume;
detecting whether the token is mapped to a snapshot pair or to a single snapshot;
when the token is mapped to a snapshot pair: copying the one or more data items from the first volume to the second volume, and transmitting to the target system a second write-via-token command that instructs the target system to copy to the one or more data items from the replica of the first volume to a replica of the second volume; and
when the token is mapped to a single snapshot: copying the one or more data items from the first volume to the second volume; and transmitting to the target system one or more write-via-hash commands that instruct the target system to copy to the one or more data items from the replica of the first volume to the replica of the second volume.