Co-residency detection in a cloud-based system
At least one virtual machine implemented on a given physical machine in an information processing system is able to detect the presence of one or more other virtual machines that are also co-resident on that same physical machine. More particularly, at least one virtual machine is configured to avoid usage of a selected portion of a memory resource of the physical machine for a period of time, and to monitor the selected portion of the memory resource for activity during the period of time. Detection of a sufficient level of such activity indicates that the physical machine is also being shared by at least one other virtual machine. The memory resource of the physical machine may comprise, for example, a cache memory, and the selected portion of the memory resource may comprise one or more randomly selected sets of the cache memory.
1. A method comprising:
configuring at least one virtual machine implemented on a physical machine to avoid usage of a selected portion of a memory resource of the physical machine for a period of time;
monitoring the selected portion of the memory resource for activity during the period of time; and
if a sufficient level of said activity is detected, generating an indication that the physical machine is also being shared by at least one other virtual machine.
2. The method of claim 1 wherein configuring the at least one virtual machine implemented on the physical machine to avoid usage of the selected portion of the memory resource of the physical machine comprises coordinating avoidance of the selected portion of the memory resource of the physical machine for the period of time by a plurality of virtual machines implemented on the physical machine.
3. The method of claim 2 wherein coordinating avoidance by the plurality of virtual machines further comprises randomly selecting a plurality of sets of a cache memory of the physical machine for a given detection period and coordinating avoidance of the plurality of sets for the detection period.
4. The method of claim 3 wherein coordinating avoidance by the plurality of virtual machines further comprises:
priming the plurality of sets of the cache memory by moving data from a main memory into the plurality of sets; and
suspending further usage of the plurality of sets by each of the plurality of virtual machines for a designated waiting interval.
5. The method of claim 4 wherein monitoring the selected portion of the memory resource for activity further comprises probing the plurality of sets after the designated waiting interval of suspended usage to determine a manner in which data contents of the plurality of sets have been altered during that interval.
6. The method of claim 5 wherein the priming, suspending and probing are repeated for each of a plurality of trials within the detection period, using the plurality of sets.
7. The method of claim 6 wherein results of said probing over multiple ones of the trials are processed to determine a classification result for the detection period indicating whether or not a level of detected activity is sufficient to indicate that the physical machine is also being shared by said at least one other virtual machine.
8. The method of claim 5 wherein probing the plurality of sets further comprises averaging cache access timing measurements over the sets.
9. The method of claim 2 wherein the coordinating and monitoring are implemented collectively by the plurality of virtual machines without requiring any separate control of an associated hypervisor.
10. The method of claim 2 wherein the plurality of virtual machines are associated with a given tenant of a cloud service provider and said at least one other virtual machine determined to be sharing the physical machine based on the detected activity is not associated with the given tenant of the cloud service provider but is instead associated with a different tenant of the cloud service provider.
11. The method of claim 1 wherein the memory resource of the physical machine comprises a cache memory.
12. The method of claim 11 wherein the cache memory comprises an L2 cache memory.
13. The method of claim 11 wherein the selected portion of the memory resource of the physical machine comprises one or more sets of the cache memory.
14. The method of claim 1 wherein the selected portion of the memory resource of the physical machine comprises a randomly selected portion of the memory resource.
15. The method of claim 1 wherein the indication is made accessible to a tenant of a cloud service provider for use in verifying exclusive use of the physical machine by the tenant.
16. The method of claim 1 further comprising:
determining an average probe timing for the period of time;
if the average is less than a specified threshold, generating a foe-absent indication; and
if the average is at or above the specified threshold, generating a foe-present indication.
17. The method of claim 1 wherein the indication comprises a classification result.
18. The method of claim 17 wherein the classification result is determined using a plurality of thresholds, and wherein each threshold is associated with a different distribution model.
19. A computer program product comprising a non-transitory processor-readable storage medium having encoded therein executable code of one or more software programs, wherein the one or more software programs when executed by a processing platform comprising the physical machine causes the processing platform to perform the method of claim 1 .
20. An apparatus comprising:
a physical machine; and
at least one virtual machine implemented on the physical machine;
wherein the at least one virtual machine is configured:
to avoid usage of a selected portion of a memory resource of the physical machine for a period of time;
to monitor the selected portion of the memory resource for activity during the period of time; and
if a sufficient level of said activity is detected, to generate an indication that the physical machine is also being shared by at least one other virtual machine.
21. The apparatus of claim 20 wherein a plurality of virtual machines implemented on the physical machine are configured to coordinate avoidance of the selected portion of the memory resource of the physical machine for the period of time.
22. The apparatus of claim 21 wherein the plurality of virtual machines are configured to coordinate avoidance of the selected portion of the memory resource of the physical machine for the period of time by coordinating avoidance of randomly selected sets of a cache memory of the physical machine for a given detection period.
23. The apparatus of claim 22 wherein the virtual machine configured to monitor the selected portion of the memory resource for activity is further configured to probe the randomly selected sets after a designated waiting interval of suspended usage to determine a manner in which data contents of the randomly selected sets have been altered during the designated waiting interval.
24. The apparatus of claim 23 wherein probing the randomly selected sets further comprises averaging cache access timing measurements over the randomly selected sets.