Flexible associativity in multitenant clustered environments
Disclosed herein are methods, systems, and processes for determining a rigidity associated with a dependent application, wherein the rigidity can be a hard rigidity, a firm rigidity, or a soft rigidity, and wherein these rigidities refer to the dependency between the dependent application and a dependee application. In response to detecting a failure of the dependee application, the disclosure provides for performing an appropriate action with respect to the dependent application, wherein that action is based, at least in part, on the rigidity associated with the dependent application. In certain embodiments, the appropriate action can further be based on a locality parameter associated with the dependent application.
1. A method comprising:
determining a rigidity associated with a dependent application, wherein
the rigidity is one of a hard rigidity, a firm rigidity, or a soft rigidity, wherein
the hard rigidity requires the dependent application to be taken offline if a first dependee application fails,
the firm rigidity allows the dependent application to remain online and be associated with a second dependee application if the first dependee application fails, and
the soft rigidity allows the dependent application to remain online and does not require the dependent application to be associated with any other dependee application if the first dependee application fails;
detecting a failure of the first dependee application; and
in response to detection of the failure, performing an appropriate action with respect to the dependent application, wherein
the appropriate action comprises
taking the dependent application offline, if the rigidity is the hard rigidity,
associating the dependent application with the second dependee application, if the rigidity is the firm rigidity, and
permitting the dependent application to continue to operate, if the rigidity is the soft rigidity.
2. The method of claim 1 , wherein
the dependent application is dependent on data received from the first dependee application.
3. The method of claim 1 , wherein
prior to the failure, the first dependee application was executing on a first node in a cluster and the dependent application was executing on a second node in the cluster.
4. The method of claim 3 , further comprising:
determining a locality requirement associated with the dependent application, wherein
the locality requirement is determined using locality information that indicates a preference for either a local locality or a global locality,
wherein
the local locality indicates that, in the case of a failover or migration of the dependent application, preference should be given to another dependee application that is operating on the first node, and
the global locality indicates that, in the case of a failover or migration of the dependent application, the dependee application can be migrated to a different node in the cluster.
5. The method of claim 4 , wherein
the appropriate action further comprises
associating the dependent application with the another dependee application, if the locality parameter indicates the preference is the local locality, and
migrating the dependent application to the different node, if the locality parameter indicates the preference is the global locality.
6. The method of claim 1 , further comprising:
subsequent to determining that the dependent application has the hard rigidity and to taking the dependent application offline, determining that the dependee application has been brought back online; and
subsequent to determining that the dependee application has been brought back online, bringing the dependent application back online.
7. The method of claim 1 , wherein:
the first dependee application and the second dependee application are both container applications,
the dependent application is a pluggable application, and
the pluggable application is a production database.
8. The method of claim 1 , wherein:
at least one of the first dependee application, the second dependee application, or the dependent application are configured to be executed by a virtual machine.
9. A non-transitory computer-readable storage medium comprising program instructions executable to perform a method comprising:
determining a rigidity associated with a dependent application, wherein
the rigidity is one of a hard rigidity, a firm rigidity, or a soft rigidity, wherein
the hard rigidity requires the dependent application to be taken offline if a first dependee application fails,
the firm rigidity allows the dependent application to remain online and be associated with a second dependee application if the first dependee application fails, and
the soft rigidity allows the dependent application to remain online and does not require the dependent application to be associated with any other dependee application if the first dependee application fails;
detecting a failure of the first dependee application; and
in response to detection of the failure, performing an appropriate action with respect to the dependent application, wherein
the appropriate action comprises
taking the dependent application offline, if the rigidity is the hard rigidity,
associating the dependent application with the second dependee application, if the rigidity is the firm rigidity, and
permitting the dependent application to continue to operate, if the rigidity is the soft rigidity.
10. The non-transitory computer-readable storage medium of claim 9 , wherein
the dependent application is dependent on data received from the first dependee application.
11. The non-transitory computer-readable storage medium of claim 9 , wherein
prior to the failure, the first dependee application was executing on a first node in a cluster and the dependent application was executing on a second node in the cluster.
12. The non-transitory computer-readable storage medium of claim 11 , wherein the method further comprises:
determining a locality requirement associated with the dependent application, wherein
the locality requirement is determined using locality information that indicates a preference for either a local locality or a global locality,
wherein
the local locality indicates that, in the case of a failover or migration of the dependent application, preference should be given to another dependee application that is operating on the first node, and
the global locality indicates that, in the case of a failover or migration of the dependent application, the dependee application can be migrated to a different node in the cluster.
13. The non-transitory computer-readable storage medium of claim 12 , wherein
the appropriate action further comprises
associating the dependent application with the another dependee application, if the locality parameter indicates the preference is the local locality, and
migrating the dependent application to the different node, if the locality parameter indicates the preference is the global locality.
14. The non-transitory computer-readable storage medium of claim 9 , wherein the method further comprises:
subsequent to determining that the dependent application has the hard rigidity and to taking the dependent application offline, determining that the dependee application has been brought back online; and
subsequent to determining that the dependee application has been brought back online, bringing the dependent application back online.
15. The non-transitory computer-readable storage medium of claim 9 , wherein:
the first dependee application and the second dependee application are both container applications,
the dependent application is a pluggable application,
the pluggable application is a production database, and
at least one of the first dependee application, the second dependee application, or the dependent application are configured to be executed by a virtual machine.
16. A system comprising:
one or more processors; and
a memory coupled to the one or more processors, wherein the memory stores program instructions executable by the one or more processors to perform a method comprising:
determining a rigidity associated with a dependent application, wherein
the rigidity is one of a hard rigidity, a firm rigidity, or a soft rigidity, wherein
the hard rigidity requires the dependent application to be taken offline if a first dependee application fails,
the firm rigidity allows the dependent application to remain online and be associated with a second dependee application if the first dependee application fails, and
the soft rigidity allows the dependent application to remain online and does not require the dependent application to be associated with any other dependee application if the first dependee application fails;
detecting a failure of the first dependee application; and
in response to detection of the failure, performing an appropriate action with respect to the dependent application, wherein
the appropriate action comprises
taking the dependent application offline, if the rigidity is the hard rigidity,
associating the dependent application with the second dependee application, if the rigidity is the firm rigidity, and
permitting the dependent application to continue to operate, if the rigidity is the soft rigidity.
17. The system of claim 16 , wherein
the dependent application is dependent on data received from the first dependee application; and
prior to the failure, the first dependee application was executing on a first node in a cluster and the dependent application was executing on a second node in the cluster.
18. The system of claim 17 , wherein the method further comprises:
determining a locality requirement associated with the dependent application, wherein
the locality requirement is determined using locality information that indicates a preference for either a local locality or a global locality,
wherein
the local locality indicates that, in the case of a failover or migration of the dependent application, preference should be given to another dependee application that is operating on the first node, and
the global locality indicates that, in the case of a failover or migration of the dependent application, the dependee application can be migrated to a different node in the cluster.
19. The system of claim 18 , wherein
the appropriate action further comprises
associating the dependent application with the another dependee application, if the locality parameter indicates the preference is the local locality, and
migrating the dependent application to the different node, if the locality parameter indicates the preference is the global locality.
20. The system of claim 16 , wherein the method further comprises:
subsequent to determining that the dependent application has the hard rigidity and to taking the dependent application offline, determining that the dependee application has been brought back online; and
subsequent to determining that the dependee application has been brought back online, bringing the dependent application back online.