SSD-block aligned writes

1. A method of managing a SSD, comprising:

partitioning a logical block address space in alignment with the SSD's blocks, giving rise to a plurality of SSD-block aligned groups, each group comprises a specific sequence of logical block addresses;

determining whether a logical block address referenced in an incoming write request is a first logical block in a respective SSD-block aligned group, and if so:

opening an ongoing SSD-block aligned write session;

assigning a SSD block to the session; and

recording in the session's data an indication of which logical block address is associated with the write data that was saved to the SSD.

2. The method according to claim 1 , further in response to determining that the logical block address is the first logical block in the group, storing respective write data in a first segment of the SSD block that was assigned to the respective ongoing SSD-block aligned write session.

3. The method according to claim 1 , wherein the ongoing SSD-block aligned write session further includes an offset parameter which indicates how many logical block addresses from the respective block aligned group were written so-far to the SSD block that was assigned to the session.

4. The method according to claim 3 , wherein in case the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, determining whether the logical block address is a successor of a latest saved logical block address of the group, and if so, storing the write data in a SSD block that was assigned to the respective group, in sequence with a latest used segment of the SSD-block, and incrementing the offset parameter.

5. The method according to claim 2 , wherein in case the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, determining whether the logical block address is a successor of a latest-saved logical block address of the group, and if so, storing the write data in a SSD block that was assigned to the respective group, in sequence with a latest used segment of the SSD-block.

6. The method according to claim 5 , wherein when write data that is associated with an entire SSD-block aligned group is stored in the SSD-block, terminating the respective ongoing SSD-block aligned write session.

7. The method according to claim 6 , wherein when following terminating an ongoing SSD-block aligned write session that was associated with a certain SSD-block aligned group, a write request referencing a logical block address that is the first logical block in the same SSD-block aligned group is received at the SSD:

opening a new ongoing write session;

assigning a new SSD block to the new session; and

recording the logical block address and the newly assigned SSD block in the new session's data.

8. The method according to claim 5 , wherein the ongoing SSD-block aligned write session further includes a free flag, and wherein when the session is opened the free flag is set to false, and when the session is terminated resetting the free flag to true.

9. The method according to claim 5 , wherein in case it is determined that the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, and is not the successor of the current last-saved logical block address of the group:

storing respective write data in a protected write buffer;

acknowledging the write request; and

determining, after a timeout period, whether the logical block address referenced in the write request that is associated with the write data that was stored in the protected write buffer is now the successor of the current latest-saved logical block address of the group.

10. The method according to claim 5 , wherein in case it is determined that the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, and is not the successor of the current last-saved logical block address of the group, storing respective write data in a SSD block that is reserved for out-of-sequence writes.

11. The method according to claim 5 , wherein in case it is determined that logical block address referenced in the incoming write request is mapped to one or more bad pages in the SSD-block assigned to the write session, storing respective write data in a SSD block that is reserved for out-of-sequence writes.

12. A SSD, comprising:

a mapping module capable of mapping each one of a plurality SSD-block aligned groups to a respective SSD block from a respective plurality of SSD blocks, wherein a SSD-block aligned group comprises a specific sequence of logical block addresses;

a controller capable of determining whether a logical block address referenced in an incoming write request is a first logical block in a respective SSD-block aligned group, and if it is, the controller is capable of:

opening an ongoing write session;

assigning a SSD block to the session; and

recording in the session's data an indication of which logical block address is associated with the write data that was saved to the SSD.

13. The SSD according to claim 12 , wherein further in response to determining that the logical block address is the first logical block in the group, the controller is configured to store respective write data in a first segment of the SSD block that was assigned to the respective ongoing SSD-block aligned write session.

14. The SSD according to claim 13 , wherein in case the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, the controller is capable of determining whether the logical block address is a successor of a latest-saved logical block address of the group, and if it is, the controller is capable of storing the write data in a SSD block that was assigned to the respective group, in sequence with a latest used segment of the SSD-block.

15. The SSD according to claim 14 , wherein when write data that is associated with an entire SSD-block aligned group is stored in the SSD-block, the controller is capable of terminating the respective ongoing SSD-block aligned write session.

16. The SSD according to claim 14 , wherein the controller is capable of including in the ongoing SSD-block aligned write session's data a free flag, and wherein when the session is opened the controller is configured to set the free flag to a false state, and when the session is terminated the controller is configured to reset the state of the free flag to true.

17. The SSD according to claim 16 , wherein when following terminating an ongoing SSD-block aligned write session that was associated with a certain SSD-block aligned group, a write request referencing a logical block address that is the first logical block in the same SSD-block aligned group is received at the SSD, the controller is capable of:

opening a new ongoing write session;

assigning a new SSD block to the new session; and

recording the logical block address and the newly assigned SSD block in the new session's data.

18. The SSD according to claim 14 , wherein in case it is determined that the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, and is not the successor of the current last-saved logical block address of the group, the controller is capable of:

storing respective write data in a protected write buffer;

acknowledging the write request; and

determining, after a timeout period, whether the logical block address referenced in the write request that is associated with the write data that was stored in the protected write buffer is now the successor of the current latest-saved logical block address of the group.

19. The SSD according to claim 14 , wherein in case the controller determines that the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, and is not the successor of the current last-saved logical block address of the group, the controller is capable of storing respective write data in a SSD block that is reserved for out-of-sequence writes.

20. The SSD according to claim 14 , wherein in case the controller determines that logical block address referenced in the incoming write request is mapped to one or more bad pages in the SSD-block assigned to the write session, the controller is capable of storing respective write data in a SSD block that is reserved for out-of-sequence writes.

21. The SSD according to claim 12 , wherein the controller is further capable of recording in the session's data an offset parameter which indicates how many logical block addresses from the respective block aligned group were written so-far to the SSD block that was assigned to the session.

22. The SSD according to claim 21 , wherein in case the logical block address referenced in the incoming write request is not the first logical block in the respective SSD-block aligned group, the controller is capable of determining whether the logical block address is a successor of a latest saved logical block address of the group, and if it is, the controller is capable of storing the write data in a SSD block that was assigned to the respective group, in sequence with a latest used segment of the SSD-block, and the controller is further capable of incrementing the offset parameter.

23. A program non-transitory storage device readable by machine, tangibly embodying a program of instructions executable by the machine to perform a method of managing a SSD, comprising:

partitioning a logical block address space in alignment with the SSD's blocks, giving rise to a plurality of SSD-block aligned groups, each group comprises a specific sequence of logical block addresses;

determining whether a logical block address referenced in an incoming write request is a first logical block in a respective SSD-block aligned group, and if so:

opening an ongoing SSD-block aligned write session;

assigning a SSD block to the session; and

recording in the session's data an indication of which logical block address is associated with the write data that was saved to the SSD.

24. A computer program product comprising a computer useable non-transitory medium having computer readable program code embodied therein of managing a SSD, the computer program product comprising:

computer readable program code for causing the computer to partition a logical block address space in alignment with the SSD's blocks, giving rise to a plurality of SSD-block aligned groups, each group comprises a specific sequence of logical block addresses;

computer readable program code for causing the computer to:

determine whether a logical block address referenced in an incoming write request is a first logical block in a respective SSD-block aligned group, and if so:

to open an ongoing SSD-block aligned write session;

to assign a SSD block to the session; and

to record in the session's data an indication of which logical block address is associated with the write data that was saved to the SSD.