Fix several oximeter bugs
#7126
Merged
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
- Only update producer information if it's actually changed. This fixes #7120, since we no longer indefinitely delay a collection timer just by refreshing our list of producers from Nexus. - Bail from a producer refresh on any error contacting Nexus. This should fix #7124. If we get a progenitor-level error refreshing our list from Nexus, which should mostly be a communcation error like `ECONNREFUSED`, we don't want to blow away our entire state. Instead, break out of the refresh function entirely, and continue collecting from our last known-good set of producers. Only proceed with the update when we are confident we have a full, valid list from Nexus. - Change the missed tick behavior for refreshing producer list from "burst" to "skip". This isn't strictly necessary, but better reflects what we want, which is "reasonably often" updates, without causing undue churn.
jgallagher
approved these changes
Nov 21, 2024
This was referenced Dec 2, 2024
jgallagher
added a commit
that referenced
this pull request
Dec 2, 2024
…ls (#7191) #7126 introduced a `return;` inside `refresh_producer_list` to avoid clobbering our producer list on an error talking to Nexus, but `refresh_producer_list` had _two_ loops: it was both "do one refresh" (from which `return`ing is correct) and also "periodically refresh" (from which `return`ing is incorrect: it causes us to never refresh again). This PR splits `refresh_producer_list` into `refresh_producer_list_{task,once}`; strongly recommend looking at the diff with whitespace ignored, as it's basically a no-op other than this split (which makes the `return` correct, as it only terminates a single refresh and not all future ones). I also added some `InlineErrorChain` bits to try to get more info from logged errors. Fixes #7190.
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.
sled_data_linkno longer returns any metrics #7120, since we no longer indefinitely delay a collection timer just by refreshing our list of producers from Nexus.oximeteris too eager to delete its entire state #7124. If we get a progenitor-level error refreshing our list from Nexus, which should mostly be a communcation error likeECONNREFUSED, we don't want to blow away our entire state. Instead, break out of the refresh function entirely, and continue collecting from our last known-good set of producers. Only proceed with the update when we are confident we have a full, valid list from Nexus.