Managing storage commands according to input-output priorities and dependencies
Input/output operations (IOs) are issued to a storage system using request queues that are each maintained for a resource targeted by the IOs. When an IO is requested, the target resource for the IO is first identified. If a request queue is maintained for the target resource, the IO is added to the request queue and the IO is issued to the storage system as the target resource becomes available. The availability of the target resource may be determined through periodic checks or by monitoring completions of IOs issued out of the request queue.
1. A method of issuing input/output operations (IOs) to a storage system in a computer system having a plurality of concurrently running execution threads, the method comprising:
identifying a resource of the storage system targeted by first and second IOs, wherein the first IO is requested by a first execution thread and the second IO is requested by a second execution thread;
adding the first and second IOs to a request queue maintained for the resource; and
if the resource is determined to be available, issuing one of the IOs to the storage system and, if the resource is determined to be unavailable, checking the availability of the resource after some time has elapsed.
2. The method of claim 1 , wherein the availability of the resource is checked repeatedly according to a predetermined schedule.
3. The method of claim 2 , wherein the predetermined schedule defines equally spaced time intervals between checks.
4. The method of claim 2 , wherein the predetermined schedule defines time intervals that are exponentially increasing between checks.
5. The method of claim 1 , further comprising:
tracking a completion of the IOs that target the resource; and
determining that the resource has become available when an IO that targets the resource has completed.
6. The method of claim 5 , wherein one of the first and second IOs is issued to the storage system according to IO priorities assigned thereto upon determining that the resource has become available.
7. The method of claim 1 , wherein the first and second IOs are added to the request queue maintained for the resource according to IO priorities assigned to the first and second IOs.
8. The method of claim 1 , wherein the computer system has a plurality of virtual machines running therein and the first and second IOs are derived from IOs issued by the virtual machines to virtual disks thereof.
9. A non-transitory computer readable medium comprising instructions that, when executed by a processor of a computer system having a plurality of concurrently running execution threads, cause the processor to carry out a method of issuing input/output operations (IOs) to a storage system, the method comprising:
identifying a resource of the storage system targeted by first and second IOs, wherein the first IO is requested by a first execution thread and the second IO is requested by a second execution thread;
adding the first and second IOs to a request queue maintained for the resource; and
if the resource is determined to be available, issuing one of the IOs to the storage system and, if the resource is determined to be unavailable, checking the availability of the resource after some time has elapsed.
10. The non-transitory computer readable medium of claim 9 , wherein the availability of the resource is checked repeatedly according to a predetermined schedule.
11. The non-transitory computer readable medium of claim 10 , wherein the predetermined schedule defines equally spaced time intervals between checks.
12. The non-transitory computer readable medium of claim 10 , wherein the predetermined schedule defines time intervals that are exponentially increasing between checks.
13. The non-transitory computer readable medium of claim 9 , wherein the method further comprises:
tracking a completion of the IOs that target the resource; and
determining that the resource has become available when an IO that targets the resource has completed.
14. The non-transitory computer readable medium of claim 13 , wherein one of the first and second IOs is issued to the storage system according to IO priorities assigned thereto upon determining that the resource has become available.
15. The non-transitory computer readable medium of claim 9 , wherein the first and second IOs are added to the request queue maintained for the resource according to IO priorities assigned to the first and second IOs.
16. The non-transitory computer readable medium of claim 9 , wherein the computer system has a plurality of virtual machines running therein and the first and second IOs are derived from IOs issued by the virtual machines to virtual disks thereof.
17. A computer system having virtual machines executed therein and a virtualization software for managing input/output operations (IOs) issued by the virtual machines to virtual disks thereof with a plurality of concurrently running execution threads, the virtualization software being programmed to:
identify a storage resource targeted by first and second IOs, wherein the first IO is requested by a first execution thread and derived from an IO issued by one of the virtual machines, and the second IO is requested by a second execution thread and derived from an IO issued by one of the virtual machines;
add the first and second IOs to a request queue maintained for the resource; and
if the resource is determined to be available, issue one of the IOs to the storage system and, if the resource is determined to be unavailable, check the availability of the resource after some time has elapsed.
18. The computer system of claim 17 , wherein the availability of the resource is checked repeatedly according to a predetermined schedule.
19. The computer system of claim 18 , wherein the virtualization software is further programmed to:
track a completion of the IOs that target the resource; and
determine that the resource has become available when an IO that targets the resource has completed.
20. The computer system of claim 17 , wherein the first and second IOs are added to the request queue maintained for the resource according to IO priorities assigned to the first and second IOs.