LMA, elesticsearch dashboards shows alive services, but cluster was shutted down 1 day ago

[Denis Klepikov]

LMA, elesticsearch dashboards shows alive services, but cluster was shutted down 1 day ago

I deploy cluster with LMA, Elesticsearch, InfluxDB, access to it web UI

poweroff all cluster

After 1 day cluster off - dashboards still show me "OKAY" (see screenshots)

Dashboards should report that servers are inaccessible

[Simon Pasquier (simon-pasquier)]

Hello Denis! The Kibana screenshot shows "connectionfailed" in the URL so the browser failed to refresh the dashboard, note that the timepicker says "from one day ago to one day ago". For Grafana I guess this is the same issue.
Can you confirm that all nodes including the InfluxDB/Grafana and Elasticsearch/Kibana nodes are down? In that case, refreshing your browser would lead to a "server not found" error.

[Denis Klepikov (dklepikov)]

connectionFailed in this case should be a "huge red button", not much people will be look on URL

[Simon Pasquier (simon-pasquier)]

@Denis, these are Kibana/Grafana UX issues. We can't do nothing much on our side.
BTW I've got an explicit message in my browser when the connection to Elasticsearch is down (see attached screenshot). Not sure why you didn't get it.

[Denis Klepikov (dklepikov)]

Unfortunately I did not see the same message into my browser (Ubuntu 14, Crome) when cluster physically disappear.
Please open a bug for Kibana/Grafana UX if needed.

[Simon Pasquier (simon-pasquier)]

Could you provide the Chrome version?

[Denis Klepikov (dklepikov)]

Version 44.0.2403.157