Extending Oracle Enterprise Manager 10g - Article
Published by Alex Gorbachev June 22nd, 2006 in Oracle“Are you missing your favorite scripts and indicators in Oracle Enterprise Manager Grid Control? Find out how easy it is to incorporate your own metrics into the Grid Control framework. Do you manage a multiple vendors’ environment and need to monitor many different components? Learn how to create your own completely new target types in Grid Control with rich set of availability and performance metrics.”
SELECT Journal published article “Extending Oracle Enterprise Manager 10g” in Quarter 2, 2006 issue.
Now you can download it on my web site here.
Interesting if Oracle will be able to push adoption of OEM so that it becomes the tool of masses. Perhaps, it’s a pricing model that prevents OEM from being popular.
Alex,
thanks for the article.
I also have been trying to use EM as monitoring tool and as a substitute for crontab tool.
Unfortunately, I can not realize repeatable notification for User-defined metrics in OEM, let’s say every-30-mintutes e-mails about objects unable to extent.
Have you tried to incarnate such functionality or some other techiques to receive notifications every time a metric crosses a threshold?
Thanks,
Andrey Goryunov
Privet Andrey,
It does look like UDM metrics are handled differently from normal ones. I recently tried to setup a UDM which was monitoring % of online redo logs required archiving and notification via email worked pretty good to me. This was on 10.2.0.1 on HP-UX (not the most stable platform for OEM).
Did you by any chance set the number of accurences to more than 1? This might affect the metric notification. Also please check history of alert for the metric - if there was no warning/critical condition detected than you obviously wouldn’t get any notifications. Maybe it’s of not much help but at least you can rule out pure notification issues. Also in the history you can see in details what notification have been sent.
Perhaps, you misunderstood the notification mechanism - as soon as warning/critical notification condition is met, alert is triggered and notifications are sent. Now, if on next collection point you still have same warning/critical condition than notification won’t be sent again. Repetitive notofication will only be sent if previously generated warning/critical condition is cleared.
HTH,
Alex
Forgot to mention the issue that I have with UDM - for some reason I can’t add some of them to Monitoring Template. I can see one but I can’t see another one when adding. I haven’t got much to investigate it so far.
Alex privet,
thanks that you found time to reply to the comment.
I am using OEM 10.2.0.1 on Solaris 32-bit and the behaviour of notification is as you said (one notification and then nothing until severity and/or UDM is changed). The number of occurences is 1. Also I see that collection timestamp for the metric is changed every established period of time and severity is still the same (since I do not do anything to avoid errors), but there are no notifications.
I described the problem/issue to Metalink guys and was told that notifications should be sent every time it checked. Might be they update a SR after my reply that there are no repeated e-mails even for predefined metrics.
I am wondering if there is a method to change severity from warning/critical to normal using some packages. In that case, I can reset flag and receive e-mails every time interval.
Thank you for your time.
Andrey
What you see is expected behavior. Otherwise, you will see alert storm in many cases. Imagine few metrics are violated and they are checked each with interval 1 or 5 minutes so you would get several notifications many time. Imagine what’s you pager or phone is going to sing that time?
If you have different requirements, perhaps, you can look at the reports or jobs approach and send notifications/emails yourself.
so cool !
just a question :
Can you send me your plugin ? or give me a link to download it ?
Thanks in advance
Regards
Jean-Philippe
Bonjour Jean-Philippe.
Well, this article describes how to build one - you can follow it and build one. It’s simplistic - just as a demo.
However, I was presenting it during UKOUG in Birmingham last month and I included the archive of sample. There is a real plug-in - it’s a bit more advanced and even has built-in reports. In addition, it’s packaged in management plug-in archive for easy deployment.
You can find the link there - http://www.pythian.com/blogs/309/ukoug-2006-presentations-available .