Atlassian uses cookies to improve your browsing experience, perform analytics and research, and conduct advertising. Accept all cookies to indicate that you agree to our use of cookies on your device. Atlassian cookies and tracking notice, (opens new window)
Reminder: Please copy/paste the Zoom chat into the notes. If you miss it, this is saved along with the meeting recording, but having it here has benefits.
*
Kong/Keycloak and flower releases
Craig
FOLIO specific Kong that adds in a little bit - start up scripts, config, etc. Pretty static. Similar for Keycloak - add in some UI, simple plugin for SSO, also pretty static.
Should these be released like modules or like infrastructure?
complicating factor - shorter support periods than FOLIO. so might need to upgrade before a flower release goes out of support, especially with longer FOLIO release cycles
Position: treat like modules because they have custom code specific to FOLIO and may require backporting
Position: treat like Kafka, postgres, etc.
Right now not too worried about backward compatibility because using very standard stuff that doesn’t change much - very unlikely to have breaking changes. CI helps prevent breaking changes
Because the changes are so minor it might be okay to update mid-flower release
Right now our K/K releases are not bound to flower releases
Sysops perspective: more complicated if you have to upgrade for each flower, better to have it independent
Does seem to resemble infrastructure more than module
Wonder if any guidelines should be set around the features we use to make sure we don’t adopt anything that is more likely to not be backwards compatible?
If someone would want to use an advanced feature, could hosts do that on their own? Not sure
Can also be revisited if needed
Start with infrastructure approach but note in OST to review how it has gone?
Do we want sysops to upgrade K/K separately from the flower release and more frequently? Yes that is the advice, esp for security reasons
How does being independent work with the OST?
Might need a different policy/category since they change version more frequently than the other infrastructure
Policy something like: Sysops should upgrade to new version when those releases are public
Tell sysops what versions are compatible with a given flower release?
Would have to do a csp for each new release?
When a new version comes out do we deploy to bugfest?
Do we want to tie it to the csp process at all?
Doesn’t really seem right for csp
Faster moving than most other infra
We haven’t previously asked sysops to upgrade in this way
Easiest might be to go ahead with new section in OST and include statement that sysops should keep them updated
Action item: update OST to include the new section and the policy that hosts are encouraged to stay on the latest version. Keep text regarding support policies and release predictions - first draft done
Action item: Release management needs to be made aware @Jakub Skoczen
Should we indicate versions that have been tested? (also for other infra like Kafka)
New releases of these should go on snapshot and bugfest - devops should be made aware @Jason Root
@Craig McNally will talk with @Oleksii Petrenko again
-
Zoom Chat
, multiple selections available, Use left or right arrow keys to navigate selected items