Nagios Support Forum

Hi All,

Its for giving a solution for monitoring using Nagios in AWS.

Requirement :

We have two availability zones in the one VPC in one region . In each Zones, we have customized Linux boxes to be monitored.Like wise we need to monitor all the Linux boxes in other regions as well.

My plan is to set up core as the poller for each availability zones and XI as the master server. For getting all the alerts in single pane of glass, i am using Fusion server to manage all the XI servers situated in different zones of different regions.

Would this plan work out...

Can someone please help me in validating this suggestion..

Regards
Rajiv

That could work, though really as long as Fusion has a path to the hosts it can integrate with either Core or XI. Though in this case, a Core "worker" and some XI "collectors" should work just fine.

Thanks Donald for the reply

We are from same team, giving some more insight to the scenario. The solution we suggested to customer is like this, we need your help to confirm whether this is a workable one or do we need to make any changes

• 2 XI Master server, one working as primary server in AWS Availability Zone 1 and another as the XI Master fail over server in Availability zone 2
• For offloading the XI Master load, I have 1 Nagios poller for Server monitoring and another 1 for application monitoring in Availability Zone1. Like that 1 each fail over pollers for server and application in Availability zone 2.
• AWS RDS MySQL instance as the DB for the XI Master servers which is in separate servers
• Nagios Fusion at the high level to connect to each XI Master server to get the consolidated view, which connects via HTTP/HTTPS

Now the questions are like this
1. Poller Fail overs
a. Whether the Nagios XI Master can manage the fail over of Availability Zone1 pollers to Availability Zone2 Pollers if the first is unavailable. If yes, then is it via DNX concept? If not please explain and whats the minimum fail over time we can set in Nagios Master?
b. Is the failover happens in poller level via heart beat check or in Master level?
2. XI Master fail over
a. How does the Nagios XI Master I in Availability Zone1 does the fail over to Nagios XI Master fail-over server in Availability Zone2? Will the pollers also failed over during this time to Availability Zone2
3. Fusion concept
a. Is Nagios Fusion can connects only to Nagios Master XIs or can in connect to each Pollers to display the data?
4. DB compatibility
a. Whether AWS RDS instance name can be used in Nagios XI Master to connect to MySQL DB?
b. Whether AWS Aurora RDS instance works with Nagios XI

Any help on the above is much appreciated

Regards,
Lijin Sarachandran

If needed i can attach the solution design arch diagram i have created

That is a lot of questions to address, and honestly is getting a bit into consulting territory.

Broadly speaking, your setup looks like it should work. I would not at all recommend DNX as it is a dead project and we have been using mod_gearman for distributed checking, but that does not directly solve the failover problem. For DR/failover setups we have been sending people to LinBit as they handle this sort of thing daily.

Can someone help me in getting the specification in calculating numbers of pollers a master xi can handle

Not 100% sure what you mean by a poller, but a single XI server can comfortably handle about 10,000 checks in a 5-minute period, and at around 20,000 checks you start to see some performance issues. This is a very broad estimate though, and is dependent on many, many factors. Your mileage may vary, and the best advice is always to baseline and extrapolate from there.