Server Management (Insight Manager 7)
1833770 Members
2121 Online
110063 Solutions
New Discussion

Maybe a very silly question

 
Scott Lovell
Advisor

Maybe a very silly question

I've finally got CIM to a stable state and we are about to start monitoring full time in a live environment. However someone asked a question the other day that got me.


When you're monitoring, do you pay more attention to the numbers on the left (Status) or on the right (Uncleared events). And how do the events on the left side get cleared.

Any information would be appreciated. This is just for curiosity sake. Thanks all.
6 REPLIES 6
Jadrice Toussaint
Honored Contributor

Re: Maybe a very silly question

Scott -

Events on the right are actual snmp traps sent form the servers when something goes wrong. this is what you will need to be notified on. the device status on the right just tells you about the state of a device. in otherwords, if you had a hardware failure that and did not clear the integrated event log on the web management, the device status would stay as is. But when you do clear the log and mark the items as repaired, the device status would change.
Jadrice Toussaint
Honored Contributor

Re: Maybe a very silly question

So to answer your question in short, the device status will not change if there is a hadrware failure until you clear the IML log. And you can do this by login into http://servername:2301. Then immediately after the snmp status for servers query is ran, the device status in IM7 will change for that server to reflect that.
Scott Lovell
Advisor

Re: Maybe a very silly question

Thanks Jadrice once again.
Brian Wright_1
Frequent Advisor

Re: Maybe a very silly question

Well you can clear the events and/or actually delete them from the database by slecting them by clicking in the whitespace of the event(s) and choosing the appropiate action from the menu. I usually just clear the events, and I have a control task configure to delete them from the database every 30 days. This allows for a history of problems when others don't do there job ;-)... I personally pay more attention to the events, and I park one console to the 'critical devices' list. Sometimes things stay down longer then they should.
Be Patient, I'm reloading.
Larry Shaw
Frequent Advisor

Re: Maybe a very silly question

We find it necessary to watch both the events and the state of a server, though I tend to pay more attention to the State than the events.

The State will usually show a server that needs some attention or that has had a more major failure that has been corrected in some manner. That said, the State may show major or minor when an event has occured and the intgrated event log on the server has not been marked repaired, but not for all event types.

There are also some cases, such as a fan removal (or a loose fan in a hot plug socket) where events will be issued showing fan removal and loss of redundancy but the state of the server will not change.

With a lot of slow links, we sent to get a fair number of "status Change - ping event" messages during backup periods, even though there are no real problems, so we tend to treat that kind of event pretty lightly unless there are accompanying events indicating a device failure.
Rob Buxton
Honored Contributor

Re: Maybe a very silly question

I always clear the "Uncleared" Events as soon as I know the cause.

That way I generally have a clear Right Hand Screen.

On the Left Hand Screen I'm trying to resolve all H/W issues that have changed a Flag Status for a Device.

So, in answer to your question, I monitor both sides and try and keep both displaying clear, that way when something happens it is quite clearly visible.
Plus I get the e-mail notifications.