Fixes #35503 - Notify user only when new errata are added to repo #10285
Conversation
|
Issues: #35503 |
|
I like that approach, it should definitely fix the repeated issues. The only concern is how scalable it is - can the dynflow steps handle inputs/outputs with (possibly tens of) thousands of errata IDs? Is there some limitation, @adamruzicka ? |
|
I have no data to back this claim up, but it should be fine. Of course the more data you push through the slower it will be, but there is no explicit limitation. Unless we get to the ballpark of around tens of megabytes in size, it should be just fine. |
|
I think it should be ok. The table ids should be much smaller than the packages data that we used to put in the package profile upload tasks and the content view publish task. |
|
@hao-yu the approach looks nice. do you have bandwidth to finish it at the moment? |
hao-yu
force-pushed
the
fix_35503_notify_only_new_errata
branch
from
7b43fe6
to
d6c295b
Compare
hao-yu
force-pushed
the
fix_35503_notify_only_new_errata
branch
from
d6c295b
to
9edb0a1
Compare
|
Sounds great to me! |
|
1 test failure seems unrelated to this patch |
|
Tests are green now. |
|
I'm running some live tests on this in my devel environment. Will do another review when that is complete. (If anyone else is interested, I had to rebase this on latest master branch to get it to start) |
|
My sincere apologies for the delay in finishing this review. After thoroughly testing this in my development environment, I came up with a few recommendations to slightly improve the UX and mostly to make the behavior around updating the inputs more consistent (in the prior implementation, the To finish before the upcoming stabilization period, I've opened a new PR #10328 which amends the original commit (keeping a single commit and so that @hao-yu 's authorship is preserved) and also adds myself and @pmoravec as co-authors. A complete write up of the changes is included, along with my testing steps. I'm requesting to please close this PR and continue the discussion at #10328 ... thanks to all for the very helpful discussion. |
|
Closing this since #10328 has merged. |
No description provided.