Normally, the Instana Agent will check for Sensor updates every morning at 4:00 server time (this is the default configuration). It is possible to change the time at which the updates are applied, the frequency at which it checks for updates or turn this feature off (no updates will be applied). See https://docs.instana.io/quick_start/agent_configuration/#agent-versioning-and-update-management for more details.
It is also possible to pin the Agent to a specific version such that it will remain at this version until configured otherwise. This KB article will explain how to pin the Agent to a specific version.
Choosing the Version to Pin the Agent to
Every change in a sensor or agent version will be released to a repository as a complete set of versions. All releases are available at https://github.com/instana/agent-updates. Each of them get a static version number, called a “sha". The sha can be found on the commits page: https://github.com/instana/agent-updates/commits/public. Every commit includes a guid as the commit identifier (e.g. fc8a533e1d41775feef02904757cb23524c7b859). This is the sha value and can be used in the Instana Agent configuration to pin the agent.
Agent installed directly on the Host
Configuration changes should be made to the <instana-home>/etc/instana/com.instana.agent.bootstrap.AgentBootstrap.cfg file.
1) Uncomment the line:
# version = <sha-1>
2) Replace <sha-1> with the sha for the version you want to pin the agent to
Please see https://docs.instana.io/quick_start/agent_configuration/#selecting-static-versions for more information.
Agent installed as a Docker Container
When executing the command to launch the Instana Agent Docker Container, environment variables can be passed to it.
1) Modify the command you use to launch the Instana Agent Docker container and pass it:
--env="INSTANA_AGENT_UPDATES_VERSION=pin the Sensor versions to a specific SHA"
Please see https://docs.instana.io/quick_start/agent_setup/container/docker/ for more info.
Agent Installed in Kubernetes using a Daemonset Yaml
When running the agent as a daemonset in k8s, the agent runs as a container. As with the docker container instructions above, environment variables can be passed in when executing the container. In this case, the environment variables can be added to the existing section under the instana-agent container section.
1) Edit the daemonset yaml file you use to deploy the Instana Agent and locate the instana-agent container definition section. Within that section, you will find the env: section with an existing set of environment variable definitions
2) Add the following:
- name: INSTANA_AGENT_UPDATES_VERSION
value: "pin the Sensor versions to a specific SHA"