Description of problem: When all gluster nodes with a volume are shut down, UI data related to the volume is missing. The volume disappears from Volume page and in grafana are no data points. Version-Release number of selected component (if applicable): tendrl-ansible-1.6.3-4.el7rhgs.noarch tendrl-api-1.6.3-3.el7rhgs.noarch tendrl-api-httpd-1.6.3-3.el7rhgs.noarch tendrl-commons-1.6.3-6.el7rhgs.noarch tendrl-grafana-plugins-1.6.3-4.el7rhgs.noarch tendrl-grafana-selinux-1.5.4-2.el7rhgs.noarch tendrl-monitoring-integration-1.6.3-4.el7rhgs.noarch tendrl-node-agent-1.6.3-6.el7rhgs.noarch tendrl-notifier-1.6.3-3.el7rhgs.noarch tendrl-selinux-1.5.4-2.el7rhgs.noarch tendrl-ui-1.6.3-3.el7rhgs.noarch How reproducible: 100% Steps to Reproduce: 1. Import cluster with disperse volume. 2. Shut down all nodes. 3. Wait an hour. 4. Open Volumes page for given cluster. Actual results: The volume disappears from UI and there are no data related to the volume. In grafana are also no data points. Expected results: The volume is shown in UI and grafana as stopped. Additional info:
when all nodes are down TTL will delete the volume details from etcd, If all nodes are down we can't monitor the cluster. so there is no point in displaying volumes.
Based on Comment 1 I close this as NOTABUG. Doc BZ 1590317 was created.
There is no clear summary backing closing decision, or references to particular meeting.
When you monitor gluster trusted storage pool with RHGS WA, shutdown all storage nodes and wait one hour, volumes hosted on these nodes will disappear from RHGS WA ui and Grafana completely. During "RHGS-3.4.0 In-flight bug triage" meeting on 2018-06-12, Nishanth pointed out that this is not a bug, and the RHGS WA is expected to do this (see also comment 1 from Gowtham). Based on this, I'm asking Nishant to verify this quick summary, and to describe how is RHGS WA expected to behave in such situation exactly instead, and then to close the BZ.
Created attachment 1450510 [details] Empty volume list when machines are down
Created attachment 1450513 [details] Cluster list with hidden volume but volume counter unchanged
Created attachment 1450520 [details] Volume dashboard when nodes are down
Here is the expected behavior: UI: * Cluster unhealthy and all hosts marked as down * No volumes and bricks listed * events reflecting the relevant status Dashboard: * Host, volume, bricks panels reflects the relevant Counts updated(up, Down etc) * Cluster/volume/brick status updated in all screens