|Ops cockpit [message #1761270]
||Tue, 09 May 2017 12:47
| Georgi Pavlov
Registered: October 2014
We need to consider how we want to evolve our support for role of the operator.|
Currently, we provide some tools for monitoring of each dirigible instance.
Supposing that we expect dirigible to run on an elastic IaaS with multiple instances we need to have an overview of the whole system int he first place and then be able to dig deep down in a concrete instance if necessary.
The major question here is how we want to support this overview?
Via our own tools or leave this to the monitoring tools of the hosting Iaas?
Can they provide sufficiently reasonable insight into the dirigible environment or they are too generic for that?
If they are good enough for drill-downs but only if we implemented vendor-specific APIs what is the best way to seamlessly deliver them?
If we are not happy with the level of support of the monitoring tools of the hosting Iaases, how our own overarching monitoring should look like? Is it a dedicated instance for monitoring only where all other agents will report?
What do we need to improve to advance our currently tools to technology cutting edge to make sure we are relevant and future proof?
Are we compliant with the trend for automation, real-time analysis, AI, and proactive measures reducing the human intervention to the bare minimum and what is missing here?
Powered by FUDForum
. Page generated in 0.02184 seconds