Back to results
US 6,738,811 B1Granted
Method and architecture for monitoring the health of servers across data networks
AssigneeSupermicro Computer, Inc.
InventorCharles Liang
Life of this patent
Estimating the term…
Abstract
A health diagnostic system capable for monitoring the health condition of computing device on a network is disclosed. According to one aspect of the present invention, a computing device is installed a module that can provoke the monitoring process that periodically samples values representing the health condition of the computing device. The sampled values are then sent back to the monitoring server for analysis. Meanwhile a monitoring server receives the sampled values and performs data analysis to determine if the computing device is in a good/poor condition based on historic data that are further used to predicate what the remaining time is for the computing device to actually break down so that necessary measures may be taken to prevent an actual breakdown.
Claims49
A method for monitoring the health of registered devices across a data network, the method comprising:
- receiving a message from the data network according to a predetermined schedule, the message including groups of data, each of the groups of data including at least one measured value reflecting current condition of a component running in one of the registered devices and surrounding values sampled substantially at the same time when the measured value is obtained;
- determining if the component is in a normal condition by comparing the measured value with a set of criteria designated for the component, wherein the criteria is based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the measured value collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration;
- and
- sending a notification to the data network when the component is determined to not be in the normal condition.
1 more element
A method for monitoring the health of registered devices across a data network, the method comprising:
- sampling a status of a number of predefined parameters of components according to a predetermined schedule;
- obtaining, substantially at the same time, respective surrounding values about each of the number of predefined parameters;
- sending a message over the network to a server, the message including the status of the predefined parameters of the components, wherein the server includes a monitoring module configured to determine whether any one of the components is in a normal condition by comparing the status of each predefined parameters with a set of criteria desigated therefor, the criteria is based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the status of the each of the predefined parameters collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration;
- and
- receiving a notification when one of the predefined parameters is determined to be abnormal.
2 more elements
A system for monitoring the health of registered devices across a data network, the system comprising:
- a number of hardware units, each installed in one of the registered devices and, when operating, measuring a status of a number of predefined parameters in one of the registered devices and obtaining, substantially at the same time, respective surrounding values about each of the number of predefined parameters;
- a number of software units, each installed in one of the registered devices and, when executing, sending a message to the data network, wherein the message includes the status as well as the respective surrounding values;
- and
- a monitoring server, coupled to the data network and upon receiving the message, providing a diagnostic analysis on the status included in the message by comparing the status of each of the predefined parameters with a set of criteria designated therefor, the criteria being based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the status of the each of the predefined parameters collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration, the monitoring server sending a notification when the diagnostic analysis determines that the status is abnormal.
1 more element
A system for monitoring the health of registered devices across a data network, the system comprising:
- a processor;
- a database for storing historic data;
- a memory, coupled to the processor, for storing code as a monitoring module;
- the code when being executed by the processor causing the monitoring module to perform operations of:
- receiving a message from the data network according to a predetermined schedule, the message including groups of data, each of the groups of data including at least one measured value reflecting current condition of a component running in one of the registered devices and surrounding values sampled at substantially the same time when the measured value is obtained;
- determining if the component is in a normal condition by comparing the measured value with a set of criteria designated for the component, wherein the criteria is based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the measured value collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration;
- and
- sending a notification to the data network when the component is determined to not be in the normal condition.
5 more elements
A computer program product running on a server computer for monitoring the health of registered devices across a data network, the computer program product comprises:
- program code for receiving a message from the data network according to a predetermined schedule, the message including groups of data, each of the groups of data including at least one measured value reflecting current condition of a component running in one of the registered devices and surrounding values sampled at substantially the same time when the measured value is obtained;
- program code for determining if the component is in a normal condition by comparing the measured value with a set of criteria designated for the component, wherein the criteria is based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the measured value collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration;
- and
- program code for sending a notification to the data network when the component is determined to not be in the normal condition.
1 more element
A computer program product running on a server computer for monitoring the health of registered devices across a data network, the computer program product comprising:
- program code for sampling a status of a number of predefined parameters of components according to a predetermined schedule;
- program code for obtaining, substantially at the same time, respective surrounding values about each of the number of predefined parameters;
- program code for sending a message over the network to a server, the message including the status of the predefined parameters of the components, wherein the server includes a monitoring module configured to determine whether any one of the components is in a normal condition by comparing the status of each of the predefined parameters with a set of criteria designated therefor, the criteria is based on at least a first and a second set of historic data, each of the first set of historic data being a consolidated representation of the status of the each of the predefined parameters collected for a specific duration, and each of the second set of historic data being a consolidated representation of one type of the surrounding values collected for the same corresponding specific duration;
- and
- program code for receiving a notification when one of the predefined parameters is determined to be abnormal.
2 more elements
Dependencies are inferred from the claim text.