Hi All
I have a general question.
We are using the Nagios Log Server appliance 1.4.0 with 2 instances.
We would like to have an IP FailOver with vrrp.
Has anybody done that with Keepalived in combination with the Nagios Log Server appliance?
Thank you in advance for your feedback.
Cheers,
Keepalived on Nagios Log Server appliance
Re: Keepalived on Nagios Log Server appliance
VRRP is not something we can directly support, as it becomes a modified system, and can create additional issues. We have not tested anything in house at this point.
With that said, I had a customer that setup VRRP not to long ago, and sent over this link for reference when I asked, so it may help you. http://everythingshouldbevirtual.com/hi ... bana-setup
With that said, I had a customer that setup VRRP not to long ago, and sent over this link for reference when I asked, so it may help you. http://everythingshouldbevirtual.com/hi ... bana-setup
Former Nagios Employee
-
nagiosnl_jorgen
- Posts: 6
- Joined: Tue Apr 22, 2014 8:39 am
Re: Keepalived on Nagios Log Server appliance
@comfone,
How may I help you getting started using keepalived? We use VRRP/keepalived on customer implementations of 2-node Nagios Log Server clusters. The main reason is that a poor man's load balancing solution like round robin DNS works fine as long as no node becomes unavailable.
To all reader of this post who might want to give Keepalived a go.. RPM packages can be found in CentOS's @base repository and rhel-7-server-rpms channel. You can configure an extra IP address (and add additional DNS A record for it) that floats between 2 nodes (or more).
The only drawback that I found is NLS sometimes shows the cluster address/name (instead of the node address/name that was elected master) under "Administration"-> "Instance Status". Also cluster_hosts sometimes lists 3 ip adresses, apart from localhost.
Regards, Jørgen van der Meulen
How may I help you getting started using keepalived? We use VRRP/keepalived on customer implementations of 2-node Nagios Log Server clusters. The main reason is that a poor man's load balancing solution like round robin DNS works fine as long as no node becomes unavailable.
To all reader of this post who might want to give Keepalived a go.. RPM packages can be found in CentOS's @base repository and rhel-7-server-rpms channel. You can configure an extra IP address (and add additional DNS A record for it) that floats between 2 nodes (or more).
The only drawback that I found is NLS sometimes shows the cluster address/name (instead of the node address/name that was elected master) under "Administration"-> "Instance Status". Also cluster_hosts sometimes lists 3 ip adresses, apart from localhost.
Regards, Jørgen van der Meulen
