When Dfmserver prcess starts, it takes almost 2 G of the memory. Attached is a screen capture which shows dramatic fall of memory usage when i stopped the Dfmserver from process management..
Is it normal, or is there anything i need to do to avoid this... Currently, i am bound to stop the dfmserver service all the time because of this issue, as the web navigation gets almost impossible with this utilization...
Total Number of Systems [Instrumented for Connectivity/Performance]: 148 [148/14
Number of Routers: 22 [22/22]
Number of Switches: 125 [125/125]
Number of Hosts: 1 [1/0]
Number of Hubs: 0 [0/0]
Number of TerminalServers: 0 [0/0]
Number of Probes: 0 [0/0]
Total Number of Links: 613
Number of Connections: 235
Number of Cables: 212
Number of Trunk Cables: 166
Number of IPs: 634 [570/0]
Number of Ports: 3625 [366/366]
Number of Interfaces: 926 [697/273]
In the Alerts and Activities Display, each alert there is comprised of events. How many active events do you have (i.e. the sum of all events in all alerts)?
Here is what i found while investigating the issue...
sm_server.exe process keeps on eating the memory, until it reaches almost 2 gig. Before that, if opened DFM->Monitoring Console, and/or DFM->Administration Console..
As soon as the sm_server.exe reaches 2 gig of mem utlization, monitoring/administration console session is terminated automatically,(i.e it automatcially detaches itself from DFM domain).
If i try to rejoin to DFM domain, i get "
Cannot attach to Domain DFM! Domain DFM not registered with broker"
Main LMS gui gets stuck, and if i try to click on any link or try to close the window, it says "The window is busy.Closing this window may cause some problems. close anyway?"
This process (of sm_server.exe taking memory upto 2 gig is iterative, i.e. after taking 2 gig, it goes back to 50 k and so, and then keeps on increasing).
Till now, onlywork around that i have is to keep the dfm_server close at all times...
Please help me, and do illustrate as to what do i need to look at.
This could be a product of the number of events in DFM. You can see these under the Monitoring Console. A large number of events could be triggering this problem. Additionally, if you are sending traps to DFM, that could also be triggering this problem.
i am sorry, but i'm still confused..
i have attached a screen capture which shows the events that are pre-configured...please help i am working towards the right direction..if by events you mean the alarms (symptoms/compounds), then all i could do about that was to increase the polling intervals, and remove the additional devices, and reduce the overheads of settings...
one thing that i noticed, is that if i stop the DFM->Administration->trap configuration->trap forwarding,
response gets a little better,(mem utilization does not get pass 1.2 g).. Also, the response of the application itself gets better.
i am trying to use DFM notification as an alternative to syslog automated action email notification.. My concerns is to monitor (get emails of) only limited interfeces on limited devices (trunk ports of core switches, edge routers only). Please provide me a hint as to how would i be able to do that?
i am kind of new to DFM, so i would appreciate if you could provide me a link to the document stating the best practices of using DFM (specially relavant attributes), providing per interface actions based email notifications OR if you could write some text providing some basic template using which i could enable my fault/error notifications on DFM..
Also, please help if trap notification is necceassary part of it, becuase i have stopped the trap notification to avoid high memory utilization..
i am required to get this DFM running with full functionality (trap receiving/forwarding, email notificatio on per interface basis) on urgent basis.. please help me in solving this problem...
I don't have the bandwidth right now to troubleshoot this. The screenshot sent previously is not what I wanted. If this is urgent, you should open a TAC service request, and have them look at this for you.
i can't a TAC request as i don't have SASU..
About events, i have monitoring console, and there are 335 entries (class), including compounds/symptoms. First Notify date is not before 3 oct, 2007.
I have attached the screen capture, i hope this is what you want...