Details
-
Improvement
-
Resolution: Fixed
-
Major
-
7.0.0, 7.0.1, 7.0.2, 7.0.3, 7.1.0
-
1
Description
To effectively implement alerting around XDCR the status of a replication needs to be known. The ns_server REST APIs return a status of "running", "paused", "error". If this value is exposed, I can create an alert on whether or not the replication has been paused as an example a customer may wish to alert on the fact a replication has been paused for more than 1 days time.
Prometheus metrics must be a number:
xdcr_status{status="running"} 1
xdcr_status{status="paused"} 0
xdcr_status{status="error"} 0
or
xdcr_status{status="running"} 0
xdcr_status{status="paused"} 1
xdcr_status{status="error"} 2
Attachments
Issue Links
- backports to
-
MB-56938 [BP 7.2.1] - XDCR No XDCR Metrics for Status of a Replication
- Closed
- depends on
-
MB-50974 No Prometheus Metrics for XDCR Errors
- Closed
- relates to
-
MB-62092 XDCR - newPipeline type errors need to be reflected on prometheus stats
- Resolved
-
MB-62096 [BP 7.6.2] - XDCR - newPipeline type errors need to be reflected on prometheus stats
- Closed
-
MB-62097 [BP 7.2.6] - XDCR - newPipeline type errors need to be reflected on prometheus stats
- Closed