View the Sprint report on Jira from the link below:
Sprint 8
With the launch approaching, the Hub is largely in a feature freeze. Sprint 8 centred on the less glamorous tasks of security, bug hunting, user testing, and incorporating a disaster recovery strategy.
Thanks to an audit by the UOA security team a handful of vulnerabilities were identified, and these were either patched immediately to scheduled for Sprint 9.
The Standup session was led by Rad, and successfully showed the development Hub switching from the master database in the Sydney datacentre to the US backup in response to a simulated loss of connection, after the switch the US became master and began mirroring transactions back to Sydney which was now acting as backup as expected.
TBD: automated switches of the kind have the potential to run amok if their decision logic for what constitutes a disaster is too promiscuous. Following jeff's advice, now the Hub has this ability, thought needs to be put into how it will be implemented.
29/05/2017 - 10/06/2017
View the Sprint report on Jira from the link below:
Sprint 9
With the launch immanent, Sprint 9 centred on the final testing, typo hunts and message improvements. Bravely, changes to flow and features were introduced, and resulting in a nicer product that coincidentally aligned very closely with ORCID's requirements for a vendor system. The last of the vulnerabilities found by the UOA security team were patched.
The backup was configured so that failure in Master would automatically switch to the backup datacentre, but any further switches would require user intervention.
The normal sprint schedule will be interrupted with the Hub going to production in an out-of-band standup session: 2:00 pm Thursday 29th June on the Hub's Zoom channel <http://bit.ly/nz-orcid-hub>
12/06/2017 - 27/06/2017