MOM 2005 at CMU Brian Richards Windows System Administrator bdr3 cmu edu System Admin Approaches The status of MOM at CMU Andrew Domain MOM 2005 sp1 What problems do we need to solve

							MOM 2005 at CMU




Brian Richards
Windows System Administrator
bdr3@cmu.edu
System Admin Approaches
The status of MOM at CMU
Andrew Domain
MOM 2005 sp1
What problems do we need to solve?

 Uptime
 Performance
 Security
 Reliability


To achieve these we use MOM 2005 and will
  transition to Ops Manager 2007
Problem One:
Monitor Our Stuff
We want to know if something happens to
 one of our servers




 Agents on all servers - discovery
 Notification to our coverage pool
  (manual paging task)
 Staff armed with Windows Mobile devices
Problem Two:
Notify Us
We want to know these things on our
 terms.



   Not all servers are the same



Our solution: Custom Heartbeat Script
Problem Three:
Tattle For Us
We want to know if someone mucks
 around with an administrator account or
 the local administrators group.




Our solution:
 Custom security-based management pack
Problem Four:
Make Us Smarter
We want detailed information about what’s
 going on with services like SQL Server or
 Microsoft Exchange.




Everyone’s solution:
  Canned management packs
  Dell MP!
Problem Five:
Keep Records
We want to be able to review what has
 happened recently.




Issue: database bloat
Problem Six:
Give Us Perf Data
We want to know how things are
 performing over time.




Issue: MOM Reports
So how have we customized MOM?
 Extended existing management packs
 Created custom management packs
 Tried using 3rd party management packs




http://www.microsoft.com/technet/scriptcenter/t
  opics/mom/mom1.mspx
Extending MPs
   Different notification responses for
    different computer groups

   Escalation of notifications

   Storage State Monitoring
Creating Custom MPs
   Security MP

   Certificate Scan MP

   VMWare mini-MP
3rd party MPs




http://www.excsoftware.com/
Providing MOM to others


 Clearly define agent requirements
 Establish SLA for monitoring
 Train clients on available alerts
 Ensure that notifications are segregated
  by service group
So where have we had issues?
   Reporting

   Not so nimble Exchange MP

   MBSA MP

   Mobile Access

   Reliability
Lessons Learned
                      Best practices
 Database             Test environment
  requirements         Agent deployment
 Reporting            MP deployment
 Agent maintenance    Alert storms
                       Fire drills
Thanks!




Brian Richards
Windows System Administrator
bdr3@cmu.edu