[GitHub] wu-sheng commented on issue #1364: Recommended architecture for micro services

[GitBox <gi...@apache.org>]

wu-sheng commented on issue #1364: Recommended architecture for micro services
URL: https://github.com/apache/incubator-skywalking/issues/1364#issuecomment-397913746
 
 
   I can give you more details about the architecture.
   
   1. Zookeeper is for collector cluster to detect each other. So agent don't need it. Collectors in cluster mode share the same zookeeper and elasticsearch cluster.
   1. Agent always need some IPs and hosts to connect collector, right? What do you prefer to use? The IPs or hosts in your settings could be considered as service discovery service provided in cluster, agent uses this to query the real collector cluster IPs, and pick one in random. In agent settings, you could set more than one IP or hostname. Even use zookeeper in agent(of course we don't), you need the whole IPs of zookeeper. So there is no difference between set the collector IPs and set zookeeper IPs. In underlying, that is same service discovery mechanism.
   1. I don't understand what is your meaning about central apm infra. APM backends need to collect, aggregate and analysis data, so, definitely they should run in a cluster. In any way, they must run a whole cluster. You can deploy collector in each VM and sharing the  same Zookeeper and elasticsearch, but I don't recommend that.
   
   Hope this can help you to understand better.

----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on GitHub and use the
URL above to go to the specific comment.
 
For queries about this service, please contact Infrastructure at:
users@infra.apache.org


With regards,
Apache Git Services