Hi Florian, I don't know if there is a more technical description somewhere, but in my experience, the inter-node protocol on port 5665 is HTTPS traffic, with the API being chatty every minute or so to continue to establish the 'heartbeat'. On top of that, check reports and check commands will generate traffic, which all depends on your configuration. With regard to network conditions, I expect it to be identical to the performance of standard HTTPS - it needs an unbroken socket to work nicely, and has a pretty normal TCP timeout depending on your OS settings. That being said, if you are OK with check results being late, you can probably have the remote ends initiate the connection and they will just sync up when they can. The replay of state and config is pretty robust to get the agent back once it has been down for a while.
hope that helps, Dave > From: "Florian Lohoff" <f...@zz.de> > To: "Icinga User's Corner" <icinga-users@lists.icinga.org> > Sent: Saturday, 6 January, 2018 12:51:42 AM > Subject: [icinga-users] Agent protocol / Distributed Monitoring > Hi, > is there a description the protocol the icinga2 agents speak > between zones/agents etc? > How sensitive is the protocol against latency, jitter, packet loss. > How much traffic is there to be expected? > I am thinking about a distributed monitoring with a couple of > satellite links which are high latency, high packet loss, limited > availability, small bandwidth. > Flo > -- > Florian Lohoff f...@zz.de > UTF-8 Test: The 🐈 ran after a 🐁, but the 🐁 ran away
_______________________________________________ icinga-users mailing list icinga-users@lists.icinga.org https://lists.icinga.org/mailman/listinfo/icinga-users
