Open
Description
Because rule evaluation for the ruler is started via the upstream Prometheus rules manager we don't have control over it's logging. The ruler passes a Querier to the rules manager as the Queryable.
The Querier starts it's own spans and logs trace IDs via spanlogger, but it's not immediately clear to me how we could have that trace ID logged when rule evaluation fails.
In order to debug things like rule evaluation failed
in the Ruler with traces stored in some tracing backends, it would be very beneficial to have trace IDs logged.