Greetings fellow IT pros,
This is my first post in the TechNet community forums so please bare with me. I was not sure if this post should be asked as a question or a discussion, but it's length considered I chose tha latter. Now, down to business.
I'm currently working as an operations consultant for a customer that is developing a ASP.NET based application for Windows Azure. We're using the RTM version, 6.1.7724.0
of the MP. What I would like to know is if anyone has any hands-on enterprise monitoring experience with this MP? I was thinking in the lines of Best Practices, pitfalls to avoid etc. For me this MP introduces a new way of thinking monitoring, not to rely
on the SCOM agent is pretty scary I must admit. Anyway, I've done some digging around, but most of the articles are based on lab experience alone. They do answer most of my questions/problems, though some remain unanswered:
- How is discovery of new deployments handled? We deploy frequently, but we still need to be sure that our monitoring is hooked up to the newly deployed instances. Is
there a setting(s) that need to be configured accordingly?
- Does the Windows Azure Role grooming work well for you? In my experience performance counter grooming runs for a while and then returns with success. However according
to the completion details I can conclude with this: sometimes nothing is deleted, and other times some performance counter records are deleted, but not all of them (not included in the grace period). Event logs are not deleted at all, it starts up fine, but
never returns anything i.e. I never receive a 34014 event for those jobs. Maybe I have to fine tune the default settings to make it work well? Any suggestions/values I might try there? I’ve tried to override on the Azure Role Class, and on the more specific
object of the Azure Role class (per application).
- How should I approach monitoring the IIS web server running in Azure? Following up on that: Are there any plans to include IIS monitoring and IIS log collection in
future versions of the MP? What have you done in your environments to monitor IIS instances running in Azure? What about collecting and parsing IIS logs in Windows Azure?
- What’s your take on the event log alerting and monitoring? Is it real time enough? The design will cause a delay based on my understanding: First the event has
to be written to table storage and then it has to be picked up by SCOM Data Collection. Haven’t tried it out myself, but I will soon enough. Anything I should be aware of when setting up Event Log Monitoring/Alerting?
FYI: All optional monitoring has been overridden to enable CPU/Memory counters, and the Windows Azure log grooming capability. I want to add that I’m not an authoring
expert, but I know a little about using the Authoring Console.
That’s all for now, but I guess new questions will arise when I identify new scenarios that has to be monitored.
Best regards
Bjørn-Erik Løken
