[Openinfralabs] Email thread discussion on cloud monitoring
bburns at redhat.com
Wed Jun 10 17:19:52 UTC 2020
We have been discussing open cloud monitoring as part of the
weekly meetings and associated with the NERC (New England
Research Cloud) Architecture (Gitlab
We are looking to develop the user stories and requirements over
the next few weeks. Note that the discussion is general and
not limited to NERC specifics, any stories or requirements relevant to
open source cloud monitoring are fair game.
We thought that using an email thread could be beneficial for people
provide their input and ideas and to discuss and even work out user
stories and requirements.
Below are some general roles and areas of monitoring that may be needed.
only a starter not by any means meant to be complete and immutable.
Consider it food for thought and discussion.
- Need to monitor cloud usage for accounting, charging
- Need to monitor historical usage for planning for future expansion
- Needs visibility to issues (problem reports) generated by users
- Logging information that may point to emerging hardware issues
- Security and compliance information
- Revision status of hardware and software
Cloud Operators / Administrators and operator tools (such as AIOPs)
- Dashboard view of the cloud, showing utilization, workloads, issues
- Low level events that will need immediate operator attention (hardware
/ power failure, platform crashes, network outages)
- Alerts generated from ongoing operations via events, metrics and
- View of current health of cloud hardware (cpu load, memory usage /
- View of performance of various aspects (memory consumption, load
averages, swapping, networking, storage)
- View of their workloads
- View of overall cloud health and capacity
Hope this helps get the conversation going!
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the Openinfralabs