关于grid control的原理

gdpr-dba

看了这篇文章http://www.askmaclean.com/tag/grid-control，对里面的某些观点有疑问，比如：
一旦Collector 收集到metric data，它会将这些度量数据和已定义的阀值做对比，检查是否发送警告(alert waring)， 同时将这些度量信息保存到本地文件系统上($OH/sysman/emd/upload目录)。 这些文件最后通过http 或 https 协议 传送到OMS服务器的指定URL上

这里说度量数据在agent端与已定义的阀值做对比，但是我觉得应该是度量数据上传到oms才和与已定义的阀值做对比的，因为阈值的定义是存储在oms的sysman里面的(mgmt_metric_thresholds)，怎么会在agent端做比对呢，不知道我的理解对不对？

Maclean Liu(刘相兵

ODM FINDING:

Threshold Values - The format of the threshold values is not correct

Applies to:
Enterprise Manager Grid Control - Version: 10.1.0.2 to 10.2.0.4
Information in this document applies to any platform.
Symptoms
In Grid Control when you navigate to the "Metric Thresholds" screen as follows:

Grid Control > Database Instance: "DB Instance Name" > Metric and Policy Settings > Metric Thresholds
The result is the following error:
"Error: Threshold Values - The format of the threshold values is not correct."
Cause
Normally this error can occur when there is an invalid character defined in a metric threshold e.g. instead
of having a straight numeric value i.e. 1000, this value is entered as 1,000 (with the comma).

Here are the valid WARNING/CRITICAL Metric Threshold values that can be used:

Combination can be:
1. (null,null) <- (as in empty, do not spell out the word "null")
2. (0,0)
3. (warning<critical) <- (warning value must be less than the critical value)
Solution
To Debug this issue:

1. A Method that we can find out which "Metric Threshold" in particular is triggering the "The format of the threshold values is not correct" error is to go to the Host for the Database where the metric threshold is being applied & take a close look at the following agent collection file: <AGENT_HOME>/sysman/emd/collection/oracle_database_<Database_Instance_Name>.xml





In this file you can check each field & compare it with it's Column Type in the metadata file.
The Metadata file can be found at the OMS level at:
<OMS_HOME>/sysman/admin/metadata/oracle_database.xml <- Here you'll get the column type definition.

2. Run the following SQL:
select COLL_NAME || '-' || NUM_OCCURENCES || '-' || WARNING_THRESHOLD ||
'-' || CRITICAL_THRESHOLD from mgmt_metric_thresholds where TARGET_GUID in
(select target_guid from mgmt_targets where target_name='<DatabaseTargetName')

NOTE: Please enter the correct Name for your Database Target experiencing this issue within the brackets <>
Check to see where the num_occurences < 1 this can help you identify the metric experiencing the problem.

4. The most common cause of this error is where the Critical Threshold is EMPTY & there is a numeric value for the Warning Threshold e.g.

Warning Threshold 85
Critical Threshold <EMPTY NOW>

To fix, just type in for 'Critical Threshold' say 97, or any value between > Warning Threshold and <100

Maclean Liu(刘相兵

FOR EXAMPLE:

AGNET_HOME


[root@nas collection]# pwd
/w01/wls/agent/agent_inst/sysman/emd/collection

[root@nas collection]# ls
oracle_beacon_EM#20#Management#20#Beacon.xml   oracle_oms_nas#3a#4889_Management_Service.xml.bak
oracle_emd_nas#3a#3872.xml                     weblogic_j2eeserver_#2f#EMGC_GCDomain#2f#GCDomain#2f#EMGC_ADMINSERVER.xml
oracle_emd_nas#3a#3872.xml.bak                 weblogic_j2eeserver_#2f#EMGC_GCDomain#2f#GCDomain#2f#EMGC_OMS1.xml
oracle_oms_nas#3a#4889_Management_Service.xml



Metric Threshold会从OMS服务器 sync到AGNET HOST的<AGENT_HOME>/sysman/emd/collection目录下

gdpr-dba

谢谢刘大，不过我还是想不通，为什么OMS的资料库里已经有阈值的定义了，还要在agent端搞个xml文件存储呢？反正度量数据是肯定要上传到oms的，等度量数据上传到oms后再和mgmt_metric_thresholds里的值作比较不可以吗？

Maclean Liu(刘相兵

http://www.askmaclean.com/grid-c ... 90%86%E5%9B%BE.html

一旦Collector 收集到metric data，它会将这些度量数据和已定义的阀值做对比，检查是否发送警告(alert waring)

由Collector最终收集到的这些信息文件仅在满足以下任意条件时实际传送给OMS:

1) 有一条alert告警信息需要发送

2) Collector收集到的信息文件的大小超过一个预定值(默认为20MB 20480KB)， 该预定限制值由$OH/sysman/config/emd.properties中UploadFileSize参数指定。

3) 从上一次数据加载算起时间超过30分钟(默认)，该预订限制值由$OH/sysman/config/emd.properties中UploadInterval 参数指定。

从设计角度来说， 开发者希望OMS与Agent之间的同步 仅在必要时发生， 例如 在agent上 将metric和 已定义的阀值做对比 如果有必要才发送一个alert warning ，显然如果agent host不存放这些 阀值信息的话无法做到这一点。