Application aware input/output fencing
Disclosed herein are methods, systems, and processes to perform application aware input/output (I/O) fencing operations. A determination is made that an instance of an application is executing on a node that is part of a cluster. In response to the determination that the instance of the application is executing on the node, an identifier that associates the instance of the application and the node on which the instance of the application is executing is generated for the instance of the application. The identifier is installed on one or more coordination points.
1. A method comprising:
determining that an instance of an application is executing on a node, wherein
the node is part of a plurality of nodes, and
the plurality of nodes are comprised a cluster; and
in response to the determination that the instance of the application is executing on the node,
generating an identifier for the instance of the application, wherein
the identifier associates the instance of the application and the node on which the instance of the application is executing, and
installing the identifier on one or more coordination points.
2. The method of claim 1 , further comprising:
determining whether one or more instances of one or more other applications are executing on the node; and
in response to the determination that the one or more instances of the one or more other applications are executing on the node,
generating one or more other identifiers for the one or more instances of the one or more other applications, wherein
the one or more other identifiers associate each of the one or more instances of the one or more other applications and the node, and
installing the one or more other identifiers on the one or more coordination points.
3. The method of claim 2 , wherein
the identifier is a registration key,
the one or more other identifiers are one or more other registration keys, and
the registration key and the one or more other registration keys are part of a plurality of registration keys.
4. The method of claim 3 , further comprising:
generating a coordination point registration key matrix, wherein
the coordination point registration key matrix comprises the plurality of registration keys,
the plurality of registration keys are stored on the one or more coordination points; and
maintaining the coordination point registration key matrix on the one or more coordination points.
5. The method of claim 4 , wherein
the plurality of nodes are communicatively coupled to each other,
the one or more coordination points are communicatively coupled to the plurality of nodes but are not part of the cluster,
the instance of the application and the one or more other instances of the one or more other applications are part of a plurality of application instances, and
each of the plurality of application instances execute on one or more nodes of the plurality of nodes.
6. The method of claim 5 , wherein
the plurality of application instances comprise a plurality of disparate application instances, and
the plurality of disparate application instances can be used to perform a plurality of Extract, Transform, and Load (ETL) processes.
7. The method of claim 5 , further comprising:
receiving an application weight matrix, wherein
the application weight matrix comprises at least one weight assigned to each of the plurality of application instances, a total application weight, and
a total node weight; and
transmitting the application weight matrix to each node of the plurality of nodes that is communicatively coupled to the node.
8. The method of claim 5 , further comprising:
determining whether the cluster is partitioned into a plurality of network partitions; and
determining whether a split-brain condition exists in the cluster as a result of the partitioning, wherein
the split-brain condition is caused by one or more application instances of the plurality of application instances executing on the one or more nodes of the plurality of nodes.
9. The method of claim 8 , further comprising:
performing an application fencing operation to rectify the split-brain condition, wherein the application fencing operation comprises, at least in part,
accessing an application weight matrix, and
performing a partition arbitration process.
10. The method of claim 9 , further comprising:
uninstalling one or more registration keys of the one or more application instances from the one or more coordination points based on a result of the partition arbitration process and as part of performing the application fencing operation, wherein
the uninstalling causes termination of the one or more application instances instead of the one or more nodes on which the one or more application instances are executing.
11. A non-transitory computer-readable storage medium (CRM) storing program instructions executable to:
determine that an instance of an application is executing on a node, wherein
the node is part of a plurality of nodes, and
the plurality of nodes are comprised a cluster; and
in response to a determination that the instance of the application is executing on the node,
generate an identifier for the instance of the application, wherein
the identifier associates the instance of the application and the node on which the instance of the application is executing, and
install the identifier on one or more coordination points.
12. The non-transitory CRM of claim 11 , further comprising:
determining whether one or more instances of one or more other applications are executing on the node; and
in response to the determination that the one or more instances of the one or more other applications are executing on the node,
generating one or more other identifiers for the one or more instances of the one or more other applications, wherein
the one or more other identifiers associate each of the one or more instances of the one or more other applications and the node, and
installing the one or more other identifiers on the one or more coordination points.
13. The non-transitory CRM of claim 12 , wherein
the identifier is a registration key,
the one or more other identifiers are one or more other registration keys, and
the registration key and the one or more other registration keys are part of a plurality of registration keys.
14. The non-transitory CRM of claim 13 , further comprising:
generating a coordination point registration key matrix, wherein
the coordination point registration key matrix comprises the plurality of registration keys,
the plurality of registration keys are stored on the one or more coordination points; and
maintaining the coordination point registration key matrix on the one or more coordination points.
15. The non-transitory CRM of claim 14 , wherein
the plurality of nodes are communicatively coupled to each other,
the one or more coordination points are communicatively coupled to the plurality of nodes but are not part of the cluster,
the instance of the application and the one or more other instances of the one or more other applications are part of a plurality of application instances,
each of the plurality of application instances execute on one or more nodes of the plurality of nodes,
the plurality of application instances comprise a plurality of disparate application instances, and
the plurality of disparate application instances can be used to perform a plurality of Extract, Transform, and Load (ETL) processes.
16. The non-transitory CRM of claim 15 , further comprising:
receiving an application weight matrix, wherein
the application weight matrix comprises at least one weight assigned to each of the plurality of application instances, a total application weight, and
a total node weight;
transmitting the application weight matrix to each node of the plurality of nodes that is communicatively coupled to the node;
determining whether the cluster is partitioned into a plurality of network partitions;
determining whether a split-brain condition exists in the cluster as a result of the partitioning, wherein
the split-brain condition is caused by one or more application instances of the plurality of application instances executing on the one or more nodes of the plurality of nodes;
performing an application fencing operation to rectify the split-brain condition, wherein the application fencing operation comprises, at least in part,
accessing the application weight matrix, and
performing a partition arbitration process; and
uninstalling one or more registration keys of the one or more application instances from the one or more coordination points based on a result of the partition arbitration process and as part of performing the application fencing operation, wherein
the uninstalling causes termination of the one or more application instances instead of the one or more nodes on which the one or more application instances are executing.
17. 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:
determine that an instance of an application is executing on a node, wherein
the node is part of a plurality of nodes, and
the plurality of nodes are comprised a cluster; and
in response to a determination that the instance of the application is executing on the node,
generate an identifier for the instance of the application, wherein
the identifier associates the instance of the application and the node on which the instance of the application is executing, and
install the identifier on one or more coordination points.
18. The system of claim 17 , further comprising:
determining whether one or more instances of one or more other applications are executing on the node; and
in response to the determination that the one or more instances of the one or more other applications are executing on the node,
generating one or more other identifiers for the one or more instances of the one or more other applications, wherein the one or more other identifiers associate each of the one or more instances of the one or more other applications and the node, and
installing the one or more other identifiers on the one or more coordination points.
19. The system of claim 18 , wherein
the identifier is a registration key,
the one or more other identifiers are one or more other registration keys, and
the registration key and the one or more other registration keys are part of a plurality of registration keys.
20. The system of claim 19 , further comprising:
generating a coordination point registration key matrix, wherein
the coordination point registration key matrix comprises the plurality of registration keys,
the plurality of registration keys are stored on the one or more coordination points; and
maintaining the coordination point registration key matrix on the one or more coordination points.
21. The system of claim 20 , wherein
the plurality of nodes are communicatively coupled to each other,
the one or more coordination points are communicatively coupled to the plurality of nodes but are not part of the cluster,
the instance of the application and the one or more other instances of the one or more other applications are part of a plurality of application instances,
each of the plurality of application instances execute on one or more nodes of the plurality of nodes,
the plurality of application instances comprise a plurality of disparate application instances, and
the plurality of disparate application instances can be used to perform a plurality of Extract, Transform, and Load (ETL) processes.
22. The system of claim 21 , further comprising:
receiving an application weight matrix, wherein
the application weight matrix comprises at least one weight assigned to each of the plurality of application instances, a total application weight, and
a total node weight;
transmitting the application weight matrix to each node of the plurality of nodes that is communicatively coupled to the node;
determining whether the cluster is partitioned into a plurality of network partitions;
determining whether a split-brain condition exists in the cluster as a result of the partitioning, wherein
the split-brain condition is caused by one or more application instances of the plurality of application instances executing on the one or more nodes of the plurality of nodes;
performing an application fencing operation to rectify the split-brain condition, wherein the application fencing operation comprises, at least in part,
accessing the application weight matrix, and
performing a partition arbitration process; and
uninstalling one or more registration keys of the one or more application instances from the one or more coordination points based on a result of the partition arbitration process and as part of performing the application fencing operation, wherein
the uninstalling causes termination of the one or more application instances instead of the one or more nodes on which the one or more application instances are executing.