For those of you who are still not aware; the days of EM being just a nice Database monitoring tool are way behind us. Curious to know what managing the complete stack means, might have persuade you to join this session by Nicole Haba (Oracle) and Ishaq Mohammed (Pacific Gas & Electric).
Nicole starts with pointing out some of the challenges we have when trying to monitor and manage the Middleware Stack.
These include:
- Monitoring & Diagnostics
- Multiple Domain Discovery
- Out-of-Box Monitoring
- Metric Extensions
- Administration Groups & Template Collections
- Incident & Problem Management
- Business Application Management
- Middleware Diagnostics Advisor
- JVM Diagnostics
- Log Viewer
- Administration
- Credential Management
- Record WLST scripts
- System MBean Browser
- Configure Log File Settings
- Change Center
- JDBC Data Source Management
- Process Control
- Selective Tracking
- Configuration Management
- Automated Asset Directory
- Compare Configurations
- Search Configurations
- Compliance Management
- Custom Configurations
- Synchronize Configurations
- Config Change History
- Lifecycle Management
- Cloning Domains
- Patch Automation
Nicole introduces some Hints and Tips… :
Like: How can we discover multiple domains in EM? Simply create a csv file containing the specs of all domains to be discovered and use EMCLI to perform the discovery, using the csv as input parameter file.
Using the Middleware Diagnostics Advisor to identify and resolve performance issues
I must admit after conducting several EM12c Bootcamps, I am a MDA fan myself…
Enforcing standards by cloning WebLogic Domains. Exactly as we are used to working with “Golden Images” of Database Homes, the same goes for WebLogic domains.
Uploading binaries to the EM Software Library (a central location in EM to store images like OS, Database, Middleware and Applications) to allow provisioning of these to new servers and by that enforcing standardization.
Next Ishaq Mohammed gives us an insight of the challenges at Pacific Gas & Electric and the way they implemented EM12c to monitor and manage the middleware stack.
Several challenges motivated PG&E to start EM12c implementation, including:
- 20% increase of WebLogic landscape over next 12 months with same number of administrators
- Unawareness of performance problems until end users complain
- Manual deployments of software being much too time-consuming, error prone, leading to inconsistent non-standard configurations
Key Use cases included:
- Administration Groups and Template Collections – to enforce standard monitoring settings
- Middleware Diagnostics Advisor – To allow for fast root cause analysis
- JVM Diagnostics – For application to disk drilldown
- Provisioning Profiles and Cloning WebLogic Domains – to enforce standard domain deployment
Using Provisioning Profiles and Cloning of WebLogic Domains allowed for SSL-Enabled WebLogic Domains with OSB deployed as well as SSL-Enabled Weblogic Domain with WebLogic Portal deployed.
It also supports the differentiation of provisioning profiles in case of dev, test or production situations.
Customized Deployment Procedures were created to provision OSB domains and WebLogic Portal domains.
Initially the EM console has been used for launching deployment procedures, to be later replaced with EMCLI.
Ishaq concludes with some final steps finalise the project including: the creation of custom reports using BI Publisher, the patching of WebLogic Servers across multiple domains, the monitoring to Service Level Agreements and the implementation of Business Transaction Management (BTM).
I must say a very interesting customer story appreciated by many in the audience.
What's cooking at Oracle - Rob Zoeteweij 