Removal of otel-collector-metrics container
The docker deployments of SigNoz has been simplified by removing the otel-collector-metrics container. The necessary metrics are now collected by the otel-collector container itself. In the past, the otel-collector-metrics container was used to scrape span metrics from the otel-collector. This container is no longer needed and has been removed.
How to upgrade if you are using otel-collector-metrics to scrape your metrics
If you are using otel-collector-metrics to scrape your metrics, you can use the otel-collector container to scrape your metrics. You can do this by moving the scrape jobs from otel-collector-metrics to otel-collector. You can find the prometheus receiver configuration in the otel-collector here
Example
If you have a scrape job like this in your otel-collector-metrics container
...
prometheus:
config:
scrape_configs:
- job_name: 'my-job'
static_configs:
- targets: ['localhost:42069']
...
You can move it to the otel-collector container like this
...
prometheus:
config:
scrape_configs:
- job_name: otel-collector
static_configs:
- targets: ['localhost:8888']
- job_name: 'my-job'
static_configs:
- targets: ['localhost:42069']
...
In case of Docker Swarm cluster with more than one node, you will have to create new collector service with the prometheus receiver and scrape configs to ensure no redundant metrics data.
How to continue using otel-collector-metrics
This is not recommended as the otel-collector-metrics container has been removed. However, you may have a use case where you want to continue using the otel-collector-metrics container. This can create merge conflicts when you upgrade SigNoz. So it is recommended to make a copy of the deployment files before upgrading. You can use the otel-collector-metrics container from the 0.36 release to continue using it.
Still need help? Join our community Slack and we will help you out.