Systems and methods for efficiently determining the health of nodes within computer clusters
View Patent ↗A computer-implemented method for determining the healthiness of nodes within computer clusters may include (1) identifying a computer cluster that includes a plurality of nodes configured to provide substantially continuous availability of at least one application, (2) identifying at least one operating system kernel installed on at least one of the nodes, (3) configuring the operating system kernel to (a) asynchronously monitor performance of the node and (b) determine, based at least in part on the node's performance, whether the node is sufficiently healthy to execute the application, (4) receiving a notification from the operating system kernel that indicates that the node is not sufficiently healthy to execute the application, and then (5) performing at least one action configured to enable the computer cluster to provide substantially continuous availability of the application despite the unhealthy node. Various other systems, methods, and computer-readable media are also disclosed.
1. A computer-implemented method for efficiently determining the health of nodes within computer clusters, at least a portion of the method being performed by a computing device comprising at least one processor, the method comprising:
identifying a computer cluster that comprises a plurality of nodes configured to provide continuous availability of at least one application;
identifying a plurality operating system kernels installed on the plurality of nodes, each operating system kernel representing part of an operating system installed on a node within the plurality of nodes and facilitating communication between the node's hardware and software;
configuring at least one operating system kernel within the plurality of operating system kernels to:
asynchronously monitor performance of the node on which the operating system kernel is installed;
determine, based at least in part on the node's performance, whether the node is healthy to execute the application;
after configuring the operating system kernel installed on the node, receiving a notification generated by the operating system kernel in response to a change in an operating status of the node, the notification indicating that the node is not healthy to execute the application;
in response to receiving the notification, performing at least one action configured to enable the computer cluster to reduce application downtime despite the unhealthy node.
2. The method of claim 1 , further comprising:
identifying a cluster engine configured to control which of the nodes is to execute the application within the computer cluster;
identifying at least one cluster agent installed on the node that enables the cluster engine to control execution of the application by the node.
3. The method of claim 2 , wherein configuring the operating system kernel comprises using the cluster engine to configure the operating system kernel.
4. The method of claim 2 , wherein performing the action configured to enable the computer cluster to provide continuous availability of the application comprises using the cluster engine to:
receive the notification from the operating system kernel;
perform the action in response to receiving the notification.
5. The method of claim 2 , wherein configuring the operating system kernel to determine whether the node is healthy to execute the application comprises configuring the operating system kernel to:
asynchronously monitor the cluster agent's health;
determine, based at least in part on the cluster agent's health, whether the node is healthy to execute the application.
6. The method of claim 2 , wherein configuring the operating system kernel comprises:
identifying an identifier associated with the cluster agent installed on the node;
registering the cluster agent's identifier with the operating system kernel to prompt the operating system kernel to asynchronously monitor the cluster agent's health.
7. The method of claim 2 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises restarting the cluster agent in an attempt to restore the cluster agent's health.
8. The method of claim 1 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises:
preventing the unhealthy node from continuing to execute the application;
selecting at least one different node within the computer cluster to begin executing the application.
9. The method of claim 1 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises ensuring that the unhealthy node is not selected to execute the application until the node becomes healthy to execute the application.
10. The method of claim 1 , further comprising:
identifying a polling mechanism installed on the node;
configuring the polling mechanism to:
periodically monitor the node's performance by checking, at a predetermined interval of time, for an unhealthy condition in which a plurality of processes initiated on the node are waiting for at least one resource to be released by one another;
determine, based at least in part on the check for the unhealthy condition, whether the node is experiencing the unhealthy condition.
11. The method of claim 10 , further comprising:
further configuring the operating system kernel to:
monitor an amount of processes being executed by the node;
determine whether the amount of processes being executed by the node is above a predetermined threshold;
after further configuring the operating system kernel, receiving a different notification from the operating system kernel that indicates that the amount of processes being executed by the node is above the predetermined threshold;
in response to receiving the different notification, dynamically adjusting the predetermined interval of time at which the polling mechanism checks for the unhealthy condition.
12. A system for determining the healthiness of nodes within computer clusters, the system comprising:
an identification module programmed to:
identify a computer cluster that comprises a plurality of nodes configured to provide continuous availability of at least one application;
identify a plurality of operating system kernels installed on the plurality of nodes, each operating system kernel representing part of an operating system installed on a node within the plurality of nodes and facilitating communication between the node's hardware and software;
a cluster engine programmed to:
configure at least one operating system kernel within the plurality of operating system kernels to:
asynchronously monitor performance of the node on which the operating system is installed;
determine, based at least in part on the node's performance, whether the node is healthy to execute the application;
receive, after configuring the operating system kernel, a notification generated by the operating system kernel in response to a change in an operating status of the node, the notification indicating that the node is not healthy to execute the application;
perform, in response to receiving the notification, at least one action configured to enable the computer cluster to reduce application downtime despite the unhealthy node;
at least one hardware processor configured to execute the identification module and the cluster engine.
13. The system of claim 12 , wherein the identification module is programmed to identify at least one cluster agent installed on the node that enables the cluster engine to control execution of the application by the node.
14. The system of claim 13 , wherein the cluster engine is programmed to configure the operating system kernel installed on the node to:
asynchronously monitor the cluster agent's health;
determine, based at least in part on the cluster agent's health, whether the node is healthy to execute the application.
15. The system of claim 13 , wherein the cluster engine is programmed to:
identify an identifier associated with the cluster agent;
register the cluster agent's identifier with the operating system kernel to prompt the operating system kernel to asynchronously monitor the cluster agent's health.
16. The system of claim 13 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises restarting the cluster agent in an attempt to restore the cluster agent's health.
17. The system of claim 12 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises:
preventing the unhealthy node from continuing to execute the application;
selecting at least one different node within the computer cluster to begin executing the application.
18. The system of claim 12 , wherein the action configured to enable the computer cluster to provide continuous availability of the application comprises ensuring that the unhealthy node is not selected to execute the application until the node becomes healthy to execute the application.
19. The system of claim 12 , wherein:
the identification module is programmed to identify a polling mechanism installed on the node;
the cluster engine is programmed to:
configure the polling mechanism to:
periodically monitor the node's performance by checking, at a predetermined interval of time, for an unhealthy condition in which a plurality of processes initiated on the node are waiting for at least one resource to be released by one another;
determine, based at least in part on the check for the unhealthy condition, whether the node is experiencing the unhealthy condition further configure the operating system kernel to:
monitor an amount of processes being executed by the node;
determine whether the amount of processes being executed by the node is above a predetermined threshold;
receive, after further configuring the operating system kernel, a different notification from the operating system kernel that indicates that the amount of processes being executed by the node is above the predetermined threshold;
dynamically adjust, in response to receiving the different notification, the predetermined interval of time at which the polling mechanism checks for the unhealthy condition.
20. A non-transitory computer-readable-storage medium comprising one or more computer-executable instructions that, when executed by at least one processor of a computing device, cause the computing device to:
identify a computer cluster that comprises a plurality of nodes configured to provide continuous availability of at least one application;
identify a plurality of operating system kernels installed on the plurality of nodes, each operating system kernel representing part of an operating system installed on a node within the plurality of nodes and facilitating communication between the node's hardware and software;
configure at least one operating system kernel within the plurality of operating system kernels to:
asynchronously monitor performance of the node on which the operating system kernel is installed;
determine, based at least in part on the node's performance, whether the node is healthy to execute the application;
receive, after configuring the operating system kernel installed on the node, a notification generated by the operating system kernel in response to a change in an operating status of the node, the notification indicating that the node is not healthy to execute the application;
perform, in response to receiving the notification, at least one action configured to enable the computer cluster to reduce application downtime despite the unhealthy node.