[teiid-issues] [JBoss JIRA] (TEIIDSB-23) Document image generation options

[ https://issues.jboss.org/browse/TEIIDSB-23?page=com.atlassian.jira.plugin... ] Steven Hawkins commented on TEIIDSB-23: --------------------------------------- For convenience moving the pr comment here: * used the term datavirtualization, rather than dataintegration as integration is already a primary term for syndesis. * discovery of the pod is based upon the label (not annotation) syndesis.io/type=datavirtualization - this can of course be changed, but needs to be done on both the deployment config for our image and in the prometheus config * metric_relabel_configs is a bit tricky. what syndesis has is redundant - a single keep causes everything that does not match to be dropped. keep also makes it hard to merge keep rules as you need to merge the regex. Note that theirs uses the context type as a prefix - which does not of course match our type labels (currently cache and runtime). We'll likely need to come up with better strategy here if we are going to share a prometheus instance. * We are following syndesis conventions. The metrics have java style names, org.teiid.xxx, but . is automatically replaced with _ in the actual name - likely to make for easier regex matching. * it is possible to separate this into just the prometheus example if it's based upon a new deployment config file that references the image built by rdbms_example and overlays the export configuration via a configmap, rather than using fabric8 include -- This message was sent by Atlassian Jira (v7.12.1#712002)