2.4.4 Notes

When displaying multiple instances of bar graphs in the Trend panel, some bar graphs may not be displayed in the graph area. If you want to check the bar graph that is not displayed, do one of the following.
- Click the instance displayed in the legend to hide the object other than the target to confirmed.
- Change the display range of the dashboard so that the time to be checked fits sufficiently within the graph area.
- Change to a stacked graph in the Edit panel window.
There is no upper limit to the number of instances displayed in the Trend panel. If there are many instances and it is difficult to see the graph, please consider defining a metric for each instance by following the steps below.

1. Change the jp1_pc_trendname of the Prometheus configuration file (jpc_prometheus_server.yml) or the discovery configuration file of the Exporter of each JP1/IM - Agent.

For details of each file, see Chapter 2. Definition Files in the JP1/Integrated Management 3 - Manager Command, Definition File and API Reference.

- Example: jpc_file_sd_config_oracledb.yml (When spliting the oracledb_exporter metrics of hostA)
```
- targets:
  - ORAHOST1:ACCTDB:30001
  labels:
    jp1_pc_exporter: JPC OracleDB exporter
    jp1_pc_category: database
    jp1_pc_trendname: oracledb_exporter_hostA
    jp1_pc_remote_monitor_instance: hostA:OracleDB metric collector(OracleDB exporter)
```
2. Copy the metric definition file of the Exporter whose metric is to be splited, rename it to the metrics_jp1_pc_trendname-label-value.conf, and place it in the same path, and then add per-instance metrics to the metrics_jp1_pc_trendname-label-value.conf.

For details of metric definition files, see Chapter 2. Definition Files in the JP1/Integrated Management 3 - Manager Command, Definition File and API Reference.

- Example: metrics_oracledb_exporter_hostA.conf (When creating a metric to display tablespace utilization for instance RBS_TS0000Q only)
```
....
  {
    "name":"tablespace_used_percent_ RBS_TS0000Q ",
    "default":true,
    "promql":"oracledb_tablespace_used_percent {tablespace="RBS_TS0000Q",type="PERMANENT"} and $jp1im_TrendData_labels",
    "resource_en":{
      "category":"database_oracle",
      "label":"Tablespace Usage_ RBS_TS0000Q ",
      "description":"Shows the usage of RBS_TS0000Q tablespace. If automatic extension is ON, it is calculated based on the size considering automatic extension.",
      "unit":"%"
    },
    "resource_ja":{
      "category":"database_oracle",
      "label":"##Japanese phrases##",
      "description":"##Japanese phrases##",
      "unit":"%"
    }
  },,
....
```
3. Add the added metric definition file to the user-created definition file list definition file, and add the target of the definition file manipulation function.

For details about the user-created definition file list definition file, see User-created definition file list definition file (imdd_user_deffile_list.json) in Chapter 2. Definition Files in the JP1/Integrated Management 3 - Manager Command, Definition File and API Reference.

- Example: imdd_user_deffile_list.json (When adding a description of the metrics_oracledb_exporter_hostA.conf)
```
{
    "filelist":[
        {
        "filename": "metrics_oracledb_exporter_hostA.conf",
        "filepath": "C:\Program Files (x86)\\Hitachi\\JP1IMM \\conf\\imdd\\plugin\\jp1pccs",
        "filecategoryID": "user_def",
        "filecategoryName": "user_def",
        "updateaction": ""
        }
    ]
}
```
4. Restart the Prometheus server.

For details, see 10. Starting and Stopping JP1/IM - Agent in the JP1/Integrated Management 3 - Manager Administration Guide.

5. Create and reflect the tree information of the IM management node.
- For Windows
  
  For details about the procedure, see 1.19.3(1)(c) Creation and import of IM management node tree data (for Windows) (required) in the JP1/Integrated Management 3 - Manager Configuration Guide.
- For Linux
  
  For details about the procedure, see 2.18.5(2)(c) Creating and importing IM management node tree data (required) in the JP1/Integrated Management 3 - Manager Configuration Guide.

When the value of the metric type Counter is obtained using the PromQL function rate, irate, increase, delta and displayed with the panel type Trend, the value of the time when the value of the counter is reset (the value of the metric to be calculated is 0) is displayed as 0 for rate, irate, and increase, and the value of the minus in the case of delta is displayed. For example:

time	counter	rate (counter[2m])	irate (counter[2m])	increase (counter[2m])	delta (counter[2m])
00:01	0	0	0	0	0
00:02	60	1	1	60	60
00:03	180	2	2	120	120
00:04	0	0	0	0	-180
00:05	60	1	1	60	60

When the value of the time series data is NaN, +Inf, or -Inf, the display for each panel type is as follows. Note that when the calculation of the PromQL statement specified in promql of the metric definition file is divided by zero, the value of the time series data becomes NaN, +Inf, or -Inf.
- For Trend:
  
  Points with time values of NaN, +Inf, or -Inf are not displayed, and line graphs do not display lines or are cut off.
- For Ranking:
  
  Instances with values NaN,+Inf, or -Inf are not displayed.
- For Number or Gauge:
  
  "No Data", "Infinity", and "-Infinity" are displayed, respectively.
If you are viewing a dashboard screen with different display ranges in multiple tabs or multiple windows of the browser, clicking the Refresh button in the Dashboard Display area changes the display range to the display range of the dashboard screen that you set the display range last. In this case, set the display range again in the dashboard where you clicked the update button.

To Page Top