Description of problem: It would be nice to add to horizon live monitoring of storage (when storage is remote it's expected to have some live errors if we fail to mount/connect to it). Version-Release number of selected component (if applicable): How reproducible: Steps to Reproduce: 1. configure gluster for cinder and glance 2. run instances from images and volumes 3. shut down the gluster servers Actual results: there is no monitoring that there is something wrong with the storage until we try to run other instances Expected results: we should see alerts that the storage is not accessible Additional info:
It sounds like something that would fall more under the remit of managing infrastructure rather than self-service for end-users. Jarda, Ladislav, is that something that the TripleO UI is looking into as well?
TripleO is looking into that but the time when this will be available is hard to estimate. It's quite likely that it will not land in Juno release.
Thank you for the reply Jarda. Is there an upstream bug or blueprint we could link to for tracking this work?
I'll make sure there is one and I will paste it here.
Thank you!
Related upstream blueprint: https://blueprints.launchpad.net/horizon/+spec/node-management-ui
This is something, tripleo UI needs to solve. Monitoring of infrastructure components is nothing horizon should show to end customers.
Hello everyone! I don't see a single hint why Ironic could be involved. Also we've done quite a few rewrites of tripleo UI, so it's probably not too valid. Anyway, Ironic is not and won't be watching hardware on the user's instance. If you plan to reopen this bug, please move it to a different component. Thanks!