Hi all,
the question is very simple, is there a way to deploy/install Filebeat in thousand (25.000/50.000) of machines?
Solutions like puppet, ansible, chef or Docker are very expensive. Maybe with few machines the cost is reasonable, but when the number grows, the cost is disproportionate.
Another question, can I use heartbeat to check the status of Filebeat?
Puppet, etc. are expensive because the license is (generally) based on the number of "managed" nodes and here the number is a bit high. On the other hand, Chef for example requires an agent installed on each machine and this is not so elengant.
Currently I don't use any kind of tool, I have only to propose a solution for this problem and eventually use it. But I think it's not so easy.
I thought you mean "expensive" from a resource specific point of view and was not thinking of the commercial part of puppet I have to feeling if you have that many machines you will not some sort of automation tool anyways. As soon as you decided for one, filebeat can be handled very similar to all the other applications you will install.
Thank you Ruflin
But, any plan to integrate this kind of feature in the next versions of Beats? Maybe in the "central configuration management" (What’s brewin’ in Beats, Elastic{on} '17)?
The monitoring part is definitively on our agenda. So far we don't have any concrete plans to also have a way to update and ship a binary as this would require to be already on the nodes. And for this tool you would probably need again some automation ...
Think about that because a way to distribute (and update) the Beats (or Elastic stack) components via SSH, maybe with a standalone app, would be the killer feature and it would make really full the stack.
Apache, Apache Lucene, Apache Hadoop, Hadoop, HDFS and the yellow elephant
logo are trademarks of the
Apache Software Foundation
in the United States and/or other countries.