Techniques for managing distributed computing components

A distributed computing system, comprising:

• a plurality of computing instances;
• a plurality of health assessment applications configured to obtain corresponding health assessment data from a respective subset of the plurality of computing instances;
• a first distributed cache;
• a second distributed cache;
• and
• a software service configured to:
• determine, from the plurality of computing instances, a set of computing instances to be probed by the plurality of health assessment applications, the set of computing instances being determined based at least in part on an estimated resource consumption for obtaining respective health assessment data corresponding to the set of computing instances and an available compute resource capacity of a computing component hosting a particular health assessment application;
• obtain, from a health assessment application of the plurality of health assessment applications, health assessment data corresponding to a computing instance of the set of computing instances;
• receive a request for first collective health assessment data of a subset of computing instances of the set of computing instances, the subset of computing instances being associated with a client;
• obtain, from the first distributed cache, the first collective health assessment data corresponding to the subset of computing instances associated with the client;
• provide the first collective health assessment data for the subset of computing instances in response to the request;
• store the first collective health assessment data in the second distributed cache as stored health assessment data;
• receive a subsequent request for second collective health assessment data of the subset of computing instances associated with the client;
• obtain, from the first distributed cache, the second collective health assessment data corresponding to the subset of computing instances;
• calculate change data indicative of a difference between the first collective health assessment data and the second collective health assessment data;
• provide the change data in response to the subsequent request;
• update the stored health assessment data with the second collective health assessment data;
• and
• when the health assessment data of at least one computing node of the subset of computing instances associated with the client is determined to be unhealthy, execute at least one remedial action comprising at least one of 1) executing operations to cause the computing instance to be removed from the distributed computing environment, or 2) transmitting a notification to a computing device that indicates the at least one computing node is unhealthy.

A non-transitory computer-readable storage medium comprising executable instructions that, when executed with one or more processors of a computing device, cause the computing device to:

• determine, from a plurality of computing instances, a set of computing instances of a distributed computing environment to be probed by a plurality of health assessment applications, the set of computing instances being determined based at least in part on an estimated resource consumption for obtaining respective health assessment data corresponding to the set of computing instances and an available compute resource capacity of a computing component hosting a particular health assessment application;
• obtain, from a health assessment application of the plurality of health assessment applications, health assessment data corresponding to a computing instance of the set of computing instances;
• receive a request for first collective health assessment data of a subset of computing instances of the set of computing instances, the subset of computing instances being associated with a client;
• obtain, from a first distributed cache, the first collective health assessment data corresponding to the subset of computing instances associated with the client;
• provide the first collective health assessment data for the subset of computing instances in response to the request;
• store the first collective health assessment data in a second distributed cache as stored health assessment data;
• receive a subsequent request for second collective health assessment data of the subset of computing instances associated with the client;
• obtain, from the first distributed cache, the second collective health assessment data corresponding to the subset of computing instances;
• calculate change data indicative of a difference between the first collective health assessment data and the second collective health assessment data;
• provide the change data in response to the subsequent request;
• update the stored health assessment data with the second collective health assessment data;
• and
• when the health assessment data of at least one computing node of the subset of computing instances associated with the client is determined to be unhealthy, execute at least one remedial action comprising at least one of:
• 1) executing operations to cause the computing instance to be removed from the distributed computing environment, or 2) transmitting a notification to another computing device that indicates the at least one computing node is unhealthy.