[Update: January 17, 2019]
Page on activity functions that can be used to check the CPU time, NIC traffic, and disk I/O via a graph
The control panel provides an activity graph function for viewing server load, disk I/O, traffic, etc retroactively. Graphs that can be checked at each resource are as follows.
|サーバ||CPU time, traffic of each loaded NIC, I/O of each connected disk|
|ルータ+スイッチ||Global network connection point traffic|
|VPCルータ||Global /individual private interface traffic|
|ロードバランサ||Loaded NIC traffic|
|データベース||CPU time, memory usage amount, loaded NIC traffic, disk usage amount, disk I/O|
|NFS||Free disk space, loaded NIC traffic|
Graph data is stored for approximately 1 year. Data prior to the past year may not displayed.
The basic operation of the activity graph screen is the same for all resources.
アクティビティグラフのデータは過去1年分保存され、その任意の期間の中から1時間, 12時間, 24時間, 3日間, 7日間, 30日間のいずれかのスケールを選択して表示することが可能です。期間の変更は左上のポップアップメニューより行うことができます。
To change the display period, click the arrow button at the top left of the screen. Use the “<” button to show information prior to the current display period, “>” for future, and “>>” to display the latest information. The display period will be indicated on the right of the button.
On the right side of the graph, the maximum and average values of each item during the currently displayed period, and the total value depending on the graph, are displayed.
Move the mouse pointer in the graph to confirm detailed data for the display period.
Details of items displayed in the activity graph are as follows. Depending on the resource type, if the item is the same there will be no difference in content.
The CPU TIME graph indicates how long the server occupied the CPU. The unit for the vertical axis is millisecond (msec). Values close to the CPU usage rate will be shown although, strictly speaking, there is a difference between them. The upper limit (rough indication) for CPU TIME that can be consumed by core is 1,000 millisecond (1 second). This means that if you have contracted for a 4-core server, the upper limit of a graph will be 4,000 msec.
CPU TIME serves as a rough indication of server load grasped by the cloud system, however the reason why load was applied on the server cannot be determined using the graph alone. Therefore, when a graph fluctuates around the upper limit, customers need to confirm themselves what kind of load has been applied, and whether or not problems are arising from the inside of their respective servers.
The graph shows a data transfer volume that indicates how much data was sent (red line) or received (green line) via the network. You can add a graph by adding a NIC.
However, packets which have been destroyed by the packet filter will not reach the target server. Therefore, those packets are not displayed on this NIC graph.
You can confirm the server transfer volume. Regular confirmation will enable prediction of future traffic trends.
This graph indicates how much data was read or written on a disk. You can add a graph by adding a disk.
Disk I/O occurred in this server due to the anti-virus software running at midnight. The performance of a server may be impaired when disk I/O increases in frequency. If you feel that response from the server is not prompt enough, we recommend checking not only CPU TIME but also disk I/O.
The data transfer volume via a router + switch will be shown. Data received from the global side (In) and data sent to the global side (Out) will be displayed in a red line and a green line, respectively. A transfer volume calculated by adding up the subordinate communication amount can be confirmed so you will be able to check whether you have sufficient room for your contract bandwidth.
*In rare cases, traffic greater in volume than the contract router + switch bandwidth will be recorded on the activity graph. A high-order router will, however, make the necessary modifications so that such high-volume traffic is adjusted to the contract communication bandwidth preventing traffic exceeding contract bandwidth from flowing to the outside.
You can confirm how much data has been transferred on the global and private sides. Where several segments have been constructed, you can see which segment is driving traffic.
On the Interface (Global) screen, data sent to the global side is shown in red, and data received from the global side in green. On the Interface (Private) screen, data sent to the subordinate switch (the server side) is indicated in red, and data received from the subordinate server in green.
For each connection switch set at the interface settings screen, Interface (Private) graphs are added in the order of Interface (Private 1), Interface (Private 2), etc.