Closed rishubhjain closed 7 years ago
This spec currently tackles only the daemonisation of the monitoring-integration code. It also assumes a single instance of monitoring-integration to be responsible for all the clusters the tendrl instance manages. Long term, monitoring-integration would need to support one or more clusters per instance, and as such needs to be able to identify the clusters is is responsible for via the cluster context provided. This requires a separate spec.
Why there is no pull request?
Highlighting an important section for me:
Install and run:
monitoring-integration will be installed in server node only. It will be installed and started by tendrl-ansible as a service. There is just one instance of monitoring-integration installed across all clusters.
@GowthamShanmugam Do you mean Tendrl Server or Storage Node? I would assume that Tendrl Server is right answer here (based on the fact that in previous milestone, we run tendrl-monitoring-integration
script on Tendrl Server), but I would like to verify this.
@mbukatov Yes Monitoring-integration will be installed on Tendrl Server, where grafana and graphite are also installed
Completed
Monitoring Integration is required to be implemented as a service instead of present implementation