Facilitating exclusive local locks on a distributed file system
Facilitating exclusive local locks on a distributed file system is provided herein. An embodiment relates to a node device of a cluster of node devices. The node device can comprise a processor and a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations. The operations can comprise determining at least one resource is to be removed from a cache of the node device. The operations also can comprise implementing an exclusive local lock over a defined lock. The defined lock can indicate interest in the at least one resource by other node devices of the cluster of node devices other than the node device. Further, the operations can comprise removing the at least one resource from the cache. The other node devices are not notified of the removing.
1. A node device of a cluster of node devices, comprising:
a processor; and
a memory that stores executable instructions that, when executed by the processor, facilitate performance of operations, comprising:
determining that at least one resource is to be removed from a cache of the node device based on receipt of an explicit request for removal of the at least one resource;
implementing an exclusive local lock over a null lock with a defined contention type, wherein the null lock indicates interest in the at least one resource by other node devices of the cluster of node devices other than the node device, and wherein the defined contention type indicates that the exclusive local lock is to be treated as an exclusive lock locally and is to be treated as a shared lock by a coordinator node; and
removing the at least one resource from the cache, wherein the other node devices are not notified of the removing, wherein the cluster of node devices comprises the coordinator node, and wherein the implementing facilitates bypass of a call to the coordinator node during the removing.
2. The node device of claim 1 , wherein the removing comprises:
determining that a reference count represented by a reference lock domain is zero; and
in response to the determining that the reference count is zero, executing a function that processes the at least one resource according to a defined procedure.
3. The node device of claim 2 , wherein the operations further comprise:
upgrading the exclusive local lock based on an activation of a contention callback indicator.
4. The node device of claim 3 , wherein the contention callback indicator is triggered based on lock contention.
5. The node device of claim 2 , wherein the executing the function comprises deleting the at least one resource.
6. The node device of claim 2 , wherein the executing the function comprises writing information related to the at least one resource to a storage device.
7. The node device of claim 2 , wherein the executing the function comprises recycling the memory of the node device.
8. The node device of claim 1 , wherein the at least one resource is at least one file stored on a distributed file system.
9. The node device of claim 1 , wherein the cluster of node devices is representative of nodes of a distributed file system.
10. A method, comprising:
determining, by a node comprising a processor and part of a group of nodes, that an object is to be unlinked at the node;
implementing, by the node, an exclusive local lock over a null lock with a defined contention type indicating interest in the object, wherein, based on the defined contention type, the exclusive local lock is treated as a shared lock by a coordinator node; and
unlinking, by the node, the object at the node, wherein the implementing facilitates, during the unlinking, bypassing a call to the coordinator node, and wherein the coordinator node is included in the group of nodes.
11. The method of claim 10 , wherein the exclusive local lock is further treated as the shared lock by a cluster-wide coordinator node and is treated as an exclusive lock by an initiator of a local node.
12. The method of claim 10 , wherein a first defined strength of the exclusive local lock on a local node is similar to a second defined strength of the shared lock or a delta lock at a cluster-wide coordinator node.
13. The method of claim 10 , wherein the determining comprises clearing cache during a shutdown operation performed at the node.
14. The method of claim 10 , further comprising:
prior to the implementing, determining, by the node, that a context of an operation being performed on the object is an operation local to the node.
15. The method of claim 10 , wherein the null lock allows one or more processes to lock the object.
16. The method of claim 10 , wherein the determining that the object is to be unlinked comprises receiving a command to delete the object.
17. The method of claim 10 , wherein the determining is based on detection of inactivity of the object at the node and a determination that the object is accessible via a second node included in the group of nodes of a distributed file system.
18. A non-transitory machine-readable medium, comprising executable instructions that, when executed by a processor, facilitate performance of operations, comprising:
determining that a cache is to be released from a first node device of a group of node devices;
identifying an object for removal from the cache;
determining whether a contention callback is assigned to the object;
implementing an exclusive local lock over a null lock with a defined contention type, wherein the exclusive local lock being implemented over the null lock indicates interest in the object by at least one second node device of the group of node devices, the at least one second node device being different than the first node device, and wherein, based on the null lock having the defined contention type, the exclusive local lock is considered locally as an exclusive lock and is considered as a shared lock by a coordinator node device; and
removing the object from the cache, wherein the implementing enables bypassing a call to the coordinator node device of the group of node devices during the removing.
19. The non-transitory machine-readable medium of claim 18 , wherein the exclusive local lock is considered the shared lock by a cluster-wide coordinator node device and is treated as the exclusive lock by an initiator of a local node device.
20. The non-transitory machine-readable medium of claim 18 , wherein the determining comprises clearing the cache during a shutdown operation performed at the first node device.