Nagios Support Forum

hi,
i want install an standby nagiosxi server . when the active nagiosxi update or have any fault.
the active nagiosxi server check more than 3W+ monitor entities.

i thanks use mysql replication and rsync files to make standby be close to active.but i don't
know the whether method have any problem .

I had search some document to found :
drdb a little complicated
backup to restore take long time
use nrdp to sync perfdata ,but configure not sync .

can you provides me some advice?

You may have seen this already but in case you haven't I'd like to draw your attention to https://assets.nagios.com/downloads/nag ... ios-XI.pdf and https://www.linbit.com/en/576-ha-nagios ... -on-rhel7/ . The first merely suggestion DRDB but the second offering a detailed guide to setting up. Hopefully this helps clarify things and makes it easier to setup the environment.

cdienger wrote:You may have seen this already but in case you haven't I'd like to draw your attention to https://assets.nagios.com/downloads/nag ... ios-XI.pdf and https://www.linbit.com/en/576-ha-nagios ... -on-rhel7/ . The first merely suggestion DRDB but the second offering a detailed guide to setting up. Hopefully this helps clarify things and makes it easier to setup the environment.

thanks for you reply,
i seriously has read the the second detailed guideline , it's mainly about: program , perfdata and mysql etc.
but the guide is far away from my requirement , which cannot meet my need.
i want a standby server that can replace the primary server when the primary one is under maintenance or is failed. what's more, Don't lost any data.

you could look at a layer below nagios with something like galera or percona for SQL replication.

from there you may hit issues with dual writes, from multiple systems, but if you ran it active / passive and preserved state it should work well.

other suggestions are shipping your concerned data to another datastore of some sort, replicating your configs, and then monitoring based on an event to trigger the passive to go active.

When you need this

Olin wrote:i want a standby server that can replace the primary server when the primary one is under maintenance or is failed. what's more, Don't lost any data.

The DRDB option you already read about is the best option. Is it complicated? yes. Is there another way to meet the requirements above without introducing additional points of failure? No.

hi @scottwilkerson,
thanks for you reply , we have two site on the different city. not provides a dedicated connection.

Olin wrote:hi @scottwilkerson,
thanks for you reply , we have two site on the different city. not provides a dedicated connection.

Unfortunately, I do not know of any solution that is going to meet your requirements that will work without a dedicated connection between sites.

Hi there,

I just want to know how to make Multiarea-synchronous.
Currently, their network latency is 100ms. and there is packet loss occasionally.

I don't think there is going to be a good solution for this, outside of DRBD that was previously suggested.

Perhaps look at Nagios Fusion to cluster instances together?

As I mentioned before, DRDB is likely the best options.

Unfortunately, Nagios XI doesn't have a HA solution built-in that can handle high latency non-dedicated connections that drop packets frequently.

Really, even if you setup the DRDB setup, you still are going to be at the mercy of the connection used to synchronize the data.