[v0.9] Reduce BundleDeployment triggering on deployed resources updates #2031
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
aruiz14
force-pushed
the
trigger-rework-0.9
branch
from
cfb49e8 to
d2dc45a
Compare
aruiz14
changed the title
aruiz14
force-pushed
the
trigger-rework-0.9
branch
from
d2dc45a to
0f7b72d
Compare
|
I like the approach. This introduces two improvements:
I don't see any issue with this approach. However, there is a possibility that these extra enqueues are needed in some edge cases we don't see at the moment. So, I wouldn't add this to the stable release (0.8) |
weyfonk
reviewed
Jan 12, 2024
aruiz14
force-pushed
the
trigger-rework-0.9
branch
from
7e1ca44 to
acd6e9c
Compare
aruiz14
force-pushed
the
trigger-rework-0.9
branch
from
acd6e9c to
e42e00d
Compare
aruiz14
force-pushed
the
trigger-rework-0.9
branch
from
e42e00d to
665bfe7
Compare
weyfonk
approved these changes
Jan 17, 2024
The watcher is stopped and started again in a short amount of time, but the update in the test happens too fast for it to start watching again.
weyfonk
reviewed
Jan 17, 2024
1 task
aruiz14
added a commit
to aruiz14/fleet
that referenced
this pull request
Mar 22, 2024
…tes" This reverts PR rancher#2031
This was referenced Mar 26, 2024
Closed
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Refers to #2058
These changes aim for reducing the number of time that the BundleDeployment reconciler is executed. This happens due to the Fleet Agent configuring a watcher on every deployed resources, which will trigger a requeue of the BundleDeployment handler in the agent for every watch event. Since this handler refreshes the conditions timestamp in the BundleDeployment status, this has a direct effect on the Fleet controller as well, which will react to the update in the BundleDeployment object.
This patch consists of the following changes:
TriggerSleepdelay has been removed, since it could cause modification events happening in that time window to be missed (this was the case for the integration tests). This is connected to the next change:ADDEDwatch events are now ignored:ADDEDevent for every object, but this does not make sense for objects that were just created.MODIFIEDandDELETED, which are the ultimate interest of the trigger (detecting modified or deleted resources). If an object is deleted and later created by the user, the controller may still need to redeploy in order to achieve a stable situation.Generationfield, for those kinds that make use of it.Generationfield will be checked: based on their UID, ifGenerationdidn't change from a previous event, it will be interpreted as the spec didn't change, so will skip the trigger, effectively omitting status updates.