Cluster Server - Service monitoring -- Conditional service
I just have to keep asking, because the pressure on me to deliver it -- is great.
For Clustered servers... I only need:
- The creation of a "<Cluster Group>" Where I could add cluster server partners to.
- The choice to actively monitor a service (or harddisk partition details/WMI mb used etc, on a clustered member server --based upon a condition---
The Condition would be
---Cluster Group Monitor: Monitor checks for the instance of this service on "any" one of the cluster member services, and monitors it from this source
--- Cluster Group Dependant Service Condition: Monitors this service only if the conditional dependant service is running. (ie, I know which of my Exchange servers is active by looking to see where DHCP is running, it does not run on the inactive nodes).
(I'm sure other users have additional input)
How you could provide this, is allow me when adding a server, to be able to select whether this server is a Cluster Member.
Non Clustered servers would create the server device just as server devices are now in OpManager.
On the Cluster Member server device, I would like "conditional" - monitor options.
For Service Monitoring, (and WMI Resource monitors) when I add a monitor to a clustered server I'd be prompted if this monitor was conditional, if YES, OpManager would:
-- Look for this service to be running on *any* of the the members of this Cluster Group.:
(EXAMPLE: I'd like to be able to monitor if DHCP is running on any one of my 4 cluster members. It only runs on the active node, and I know the cluster member server where DHCP is running, also now owns the QUORUM resource --shared partition. ).
OpManager would check each clustered device and return an "active status" for the service, and additionally let me know which of the Clustered server members was running the service. (the change in location of the service could be an event notification).
I'd make use of WMI resource conditional monitoring as follows:
-- I'd like to be able to monitor the "F" drive (etc.) , in the same manner because it's the "quorum" drive, which can only be seen by the active Cluster Member server. In fact the presence of DHCP running on this server (just happens to be), tells me the partition monitoring of drive F can be done from here.
I want detailed history of the used disk space on this Quorum partition over time, so no matter which server currently "owns" it, I want to track that partition information for as long as possible.
-- WMI EXCHANGE monitoring! As it happens, that Quorum drive is where my Exchange Information Store is kept. I want all the dandy Exchange Queaue monitoring to be available to me (and so does our Exchange Administrator).
We have ten email servers. 4 of which are currently clustered (it'll go back down to 2 node cluster, after we complete our Exchange 2003 migration). Email is important, we really need some ability to monitor with OpManager our Clustered environment. I don't mind doing some extra work, troubleshooting, configuring etc., to get this to work for us.
Thank you
For Clustered servers... I only need:
- The creation of a "<Cluster Group>" Where I could add cluster server partners to.
- The choice to actively monitor a service (or harddisk partition details/WMI mb used etc, on a clustered member server --based upon a condition---
The Condition would be
---Cluster Group Monitor: Monitor checks for the instance of this service on "any" one of the cluster member services, and monitors it from this source
--- Cluster Group Dependant Service Condition: Monitors this service only if the conditional dependant service is running. (ie, I know which of my Exchange servers is active by looking to see where DHCP is running, it does not run on the inactive nodes).
(I'm sure other users have additional input)
How you could provide this, is allow me when adding a server, to be able to select whether this server is a Cluster Member.
Non Clustered servers would create the server device just as server devices are now in OpManager.
On the Cluster Member server device, I would like "conditional" - monitor options.
For Service Monitoring, (and WMI Resource monitors) when I add a monitor to a clustered server I'd be prompted if this monitor was conditional, if YES, OpManager would:
-- Look for this service to be running on *any* of the the members of this Cluster Group.:
(EXAMPLE: I'd like to be able to monitor if DHCP is running on any one of my 4 cluster members. It only runs on the active node, and I know the cluster member server where DHCP is running, also now owns the QUORUM resource --shared partition. ).
OpManager would check each clustered device and return an "active status" for the service, and additionally let me know which of the Clustered server members was running the service. (the change in location of the service could be an event notification).
I'd make use of WMI resource conditional monitoring as follows:
-- I'd like to be able to monitor the "F" drive (etc.) , in the same manner because it's the "quorum" drive, which can only be seen by the active Cluster Member server. In fact the presence of DHCP running on this server (just happens to be), tells me the partition monitoring of drive F can be done from here.
I want detailed history of the used disk space on this Quorum partition over time, so no matter which server currently "owns" it, I want to track that partition information for as long as possible.
-- WMI EXCHANGE monitoring! As it happens, that Quorum drive is where my Exchange Information Store is kept. I want all the dandy Exchange Queaue monitoring to be available to me (and so does our Exchange Administrator).
We have ten email servers. 4 of which are currently clustered (it'll go back down to 2 node cluster, after we complete our Exchange 2003 migration). Email is important, we really need some ability to monitor with OpManager our Clustered environment. I don't mind doing some extra work, troubleshooting, configuring etc., to get this to work for us.
Thank you