This is likely a "bigger" ticket, but I wanted to add this as a conversation point.
Since June 2021, we have been doing a periodic manual ontology update in minerva. This procedure currently requires two people, a pre-coordinated and publically communicated time, and around 2hrs. To date, it has been done around 64 times.
Additionally, given the manual nature and the fact that this is a high-value production system, when things do inevitably go wrong, we need to be cautious, sometimes leading to longer outages, redos, and general stress; e.g.:
Finally, given that this is a production system, we try to do outages when things are "quite", typically meaning that we are encroaching on hours that would typically be considered "off" for the people performing these tasks.
Given all this, I would propose that automation would reduce errors and time spent, at least in the long-run. There are two approaches:
have minerva do migrations and updates automatically and internally
using automation scripts, create a turnkey system where a newly migrated and updated system is brought up (and automatically tested)
This is likely a "bigger" ticket, but I wanted to add this as a conversation point.
Since June 2021, we have been doing a periodic manual ontology update in minerva. This procedure currently requires two people, a pre-coordinated and publically communicated time, and around 2hrs. To date, it has been done around 64 times.
Additionally, given the manual nature and the fact that this is a high-value production system, when things do inevitably go wrong, we need to be cautious, sometimes leading to longer outages, redos, and general stress; e.g.:
Finally, given that this is a production system, we try to do outages when things are "quite", typically meaning that we are encroaching on hours that would typically be considered "off" for the people performing these tasks.
Given all this, I would propose that automation would reduce errors and time spent, at least in the long-run. There are two approaches:
Tagging @vanaukenk