Drop using try / catch directly and instead,

update handleFail fn to not slack notifiy
if the build is unstable, per Brian S.

.ci/Jenkinsfile_coverage

@@ -11,13 +11,11 @@ kibanaPipeline(timeoutMinutes: 240) {
  'CODE_COVERAGE=1', // Enables coverage. Needed for multiple ci scripts, such as remote.ts, test/scripts/*.sh, schema.js, etc.
  ]) {
  workers.base(name: 'coverage-worker', size: 'l', ramDisk: false, bootstrapped: false) {
- try {
+ catchError {
  kibanaCoverage.runTests()
  handleIngestion(TIME_STAMP)
- } catch (err) {
- currentBuild.result = 'FAILURE'
- handleFail()
  }
+ handleFail()
  }
  }
  kibanaPipeline.sendMail()
@@ -36,7 +34,7 @@ def handleIngestion(timestamp) {
 
 def handleFail() {
  def buildStatus = buildUtils.getBuildStatus()
- if(params.NOTIFY_ON_FAILURE && buildStatus != 'SUCCESS' && buildStatus != 'ABORTED') {
+ if(params.NOTIFY_ON_FAILURE && buildStatus != 'SUCCESS' && buildStatus != 'ABORTED' && buildStatus != 'UNSTABLE') {
  slackNotifications.sendFailedBuild(
  channel: '#kibana-qa',
  username: 'Kibana QA'