SYSTEM MONITORING DEVICE AND SYSTEM MONITORING METHOD
To solve the problem that, in order to determine an abnormal state from operation information, it is necessary to determine an abnormality value from operation information in abnormal states that occurred in the past or states indicating signs of abnormality, and predicting the abnormality sign is u...
Gespeichert in:
Hauptverfasser: | , , |
---|---|
Format: | Patent |
Sprache: | eng ; jpn |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
Zusammenfassung: | To solve the problem that, in order to determine an abnormal state from operation information, it is necessary to determine an abnormality value from operation information in abnormal states that occurred in the past or states indicating signs of abnormality, and predicting the abnormality sign is unfeasible when there is no data that can be clearly determined as an abnormality.SOLUTION: In a system monitoring device 10, a monitoring manager 40 acquires operation information for each process in each of computers 20 and 21 collected by monitoring agents 50 and 51, automatically selects a significant process to be monitored, and calculates collection conditions for its operation information. By collecting and accumulating detailed operation information for a long period of time according to the collection conditions, it becomes possible to diagnose changes in the value of the operation information from the past in detail and detect signs leading to failure.SELECTED DRAWING: Figure 1
【課題】稼働情報から異常状態を判別するためには、過去に発生した異常状態や異常の兆候を示す状態における稼働情報から異常の値を決めておく必要があり、明確に異常と判断できるデータがないと異常兆候の予測ができない。【解決手段】システム監視装置10において、監視マネージャ40は、監視エージェント50、51が収集した各計算機20、21におけるプロセス毎の稼働情報を取得し、監視する対象とすべき重要プロセスを自動選別し、その稼働情報の収集条件を算出して、この収集条件に従って、詳細な稼働情報を長期間収集し蓄積することで、過去からの稼働情報の値の変化を詳細に診断可能となり、障害に至る兆候を検出できるようにした。【選択図】 図1 |
---|