hey there,
I have to be honest and admit that I have several doubt about Elastic Agent and Fleet so I was trying to monitor a production-fake Elasticsearch cluster using Elastic Agent enrolled by Fleet Server.
I follow this repo and its blog article.
Since I want to monitor another Elasticsearch cluster I was trying to add the proper integration but when I have to execute the following command on one node
curl -L -O https://artifacts.elastic.co/downloads/beats/elastic-agent/elastic-agent-8.16.1-linux-x86_64.tar.gz
tar xzvf elastic-agent-8.16.1-linux-x86_64.tar.gz
cd elastic-agent-8.16.1-linux-x86_64
sudo ./elastic-agent install --url=https://ecp-fleet-server:8220 --enrollment-token=xxxxxxx
I got this error:
rosetta error: failed to open elf at /lib64/ld-linux-x86-64.so.2
Trace/breakpoint trap
I cannot understand how to move on. any suggest?
The whole systems (monitoring and production-fake) are based on 8.16.1
You also need to check if you tried to do something like change your default architecture ... You should not need to.
As far as I understand Rosetta should not be involved
Did you do something like this
Close the terminal. Open the terminal reboot the Mac. Whatever you need to do, but the arm aarch64 should run on your machine assuming you have not modified it in some way.
Wait now I reread you're trying to use the agent with docker?
Is that correct?
What exactly are you trying to do run the agent as a binary install or as a docker container.
I am still on this issue, even if I restart everything, destroyed the original container (and volume) and created it again.
Anyway, inside the elasticsearch container (where I should install the Elastic Agent with my Elasticsearch policy) I cannot use sudo command. Really tricky since not every step is really clear.
Can you give more context on what you are trying to do, which commands you are running and where?
You do not install an Elastic Agent inside the Elasticsearch container.
If you want to run Elastic Agent as docker container, you do not install it, you need to follow this documentation to run an Agent container.
If you want this agent to get logs from your Elasticsearch containers, then you would need to configure the Elasticsearch containers to save the logs on a persistent bind volume and also mount this same volume inside the Agent container.
I have a monitoring cluster and a production cluster. Every cluster uses container technology.
I want to monitor the production cluster using Elastic Agent, so I have started an Elastic Agent with the Fleet Server inside enabled. Then I have configured Fleet Server using Kibana UI and created an Integration (Elasticsearch).
Then, I have to add an Elastic Agent, so I have to execute the commands posted 3-4 messages above.
I am following these articles:
this one, in theory I could also skip it and use the immortal elastic:xxx
and this one to create policy/integration/agent
The monitoring stack is based on this project
What would I like to get? To monitor my production cluster with Elastic Agents, in a specific way the metrics. As far as I understood, using Elastic Agent rather than Beats should be more easy and efficient.
As @leandrojmp indicated, it is against best practice to install additional software inside of containers. Instead, we deploy an additional container running our additional software (sometimes called a sidecar).
We then configure the sidecar container to connect to the Elasticsearch container to perform monitoring.
Ok, so let me try to summarise the correct scenario.
Machine 1: elasticsearch (monitoring cluster) - kibana - elastic agent (fleet server role).
Machine 2: elasticsearch (production cluster) - elastic agent (to deploy the agent with elasticsearch integration)
Every component, both machines, are containerised.
Is this scenario correct?
Elastic Agent maintains a persistent connection to the fleet server (initiated by Elastic Agent) to receive policy updates. Almost like a persistent/continuous pull mode.
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.
