Send alarm when server is shutdown or crash.

[Miller Echagarreta]

Hi all.
I have some doubts for monitoring availability of a server.
I have a server with robot 7.97 and connect it to hub version 7.97. Right now I am see a metric "Power State" from hub, but ¿can I use this metric for send alarm when servidor has been off?. My question is because I need find a strategy for monitoring availability server without use ICMP.
Where can I configure the treshold for send alarm when this qos "Power State" has been 0

[David Michel]

It might work via Operator Console > Settings > Alarm Policy > Device > Infrastructure > 'Power State' shows up there.

------------------------------
Support Engineer
Broadcom
------------------------------

[Garin Walsh]

Monitoring the status of a server requires a second server to do the monitoring.

net_connect by default uses icmp but you can configure other TCP based services - like connecting to port 48000 for the Nimbus robot.

Alternatively you can query a QOS table (like the power state QOS) that's consistently updated and measure the time since last update. Alert on a long time.

Otherwise, ironically, there's nothing in the product that natively alerts on whether some arbitrary server is up or not.

[Miller Echagarreta]

Hi @David Michel​
We don't have UIM 902, we have 8.51 SP1, right now in our planning the upgrade CAUIM will be by early 2020.
Exist other procedure for capture alarm of "Power State", when the robot_power_state = 0
Maybe I think about to create a dashboard with this metric "Power State" and change the color when the metric is down but I have 1250 servers I don't see how this is good or best practice. Any idea?

[David Michel]

There is the robot inactive alarm

Article title: How to customize the message of the robot inactive alarm
Article Id: 135748
https://ca-broadcom.wolkenservicedesk.com/external/article?articleId=135748

Article title: What is the cause of an Alarm if a server robot goes down?
Article Id: 34254
https://ca-broadcom.wolkenservicedesk.com/external/article?articleId=34254

Article title: What is the delay for Inactive Robot Alarm
Article Id: 125149
https://ca-broadcom.wolkenservicedesk.com/external/article?articleId=125149

------------------------------
Support Engineer
Broadcom
------------------------------

[Garin Walsh]

Robot inactive only works if the hub is still functional and able to send that message. If that down robot is also the hub, then if that system goes down, you don't get a message and you don't have a good indicator.

There's a queue down message and a tunnel down message if you are using queues or tunnels to that robot/hub but these fail for their own reasons which might have nothing to do with the hub/robot being down.

[Miller Echagarreta]

Thanks to all. The solutions was to configure the monitoring nimbus service 48000 TCP port with net_connect and the customer it's evaluating to use icmp in all servers.

[Daniel Blanco]

Hi Dave,
In two of the article posted above, none of the pictures show.
Is this an issue with the site or were pictures lost?

------------------------------
Daniel Blanco
Enterprise Tools Architect
Alphaserve Technologies
------------------------------

[Kathy Maguire]

Daniel -

This is a known issue with several Knowledge Documents.  Wolkensoft support is currently actively working on getting this corrected.  October 14 is the latest ETA for having this fixed.

------------------------------
Kathy Maguire
Technical Support Engineer 4
Broadcom
------------------------------