Preemptive self-healing of application server hanging threads

A method for correcting stuck thread scenarios on an application server, comprising:

• taking, by a hardware processor communicatively coupled to a memory, snapshots of an application server's performance parameters at periodic time intervals, wherein the snapshots include the status of a plurality of performance parameters;
• determining, by the hardware processor, that a thread running on the application server is stuck when it has been working on a request for more than a configured maximum time;
• retrieving, by the hardware processor, a set of snapshots, from the memory, of the system performance taken during a period of time preceding the moment when it was determined that the thread running on the application server is stuck;
• identifying, by the hardware processor, from among the retrieved set of snapshots the oldest snapshot with a performance parameter that exceeds a threshold;
• comparing, by the hardware processor, the server performance parameters comprising the oldest snapshot with a repository of previous stuck thread scenarios, wherein the stuck thread scenario comprises the snapshot of the system performance taken when the thread has been working on a request for more than a configured maximum time, to identify a closest match
• identifying, by the hardware processor, a script with the instructions for resolving the matched stuck thread scenario from a database of solutions to the previous stuck thread scenarios;
• and
• executing, by the hardware processor, the identified script.

An apparatus for correcting stuck thread scenarios on an application server, comprising:

• a memory operable to store:
• snapshots of an application server's performance that are taken at periodic intervals, wherein the snapshots include a status of a plurality of server performance parameters;
• a repository of previous stuck thread scenarios, wherein a stuck thread scenario comprises the snapshot of the system performance taken when the thread has been working on a request for more than a configured maximum time;
• and a plurality of scripts, each comprising a set of instructions for resolving one of the stuck thread scenarios;
• a hardware processor communicatively coupled to the memory, the hardware processor configured to:
• monitor the plurality of server performance parameters;
• determine that a thread running on the application server is stuck when it has been working on a request for more than a configured maximum time;
• retrieve, from the memory, a set of snapshots of the system performance taken for a period of time preceding the moment when it was determined that the thread running on the application server is stuck;
• identify from among the retrieved set of snapshots the oldest snapshot with a server parameter that exceeds a threshold;
• compare the server performance parameters comprising the oldest snapshot with the repository of previous stuck thread scenarios to identify a closest match;
• identify a script with the instructions for resolving the matched stuck thread scenario;
• and
• execute the identified script.