introduce timeout factor to control invoker queue behavior #3767
Conversation
This file contains hidden or 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
rabbah
reviewed
Jun 15, 2018
| min = 100 ms | ||
| max = 5 m | ||
| std = 1 m | ||
| timeoutfactor = 1 |
Member
There was a problem hiding this comment.
Can you add a comment as this knob isn’t obvious.
Codecov Report
@@ Coverage Diff @@
## master #3767 +/- ##
=========================================
- Coverage 75.84% 75.14% -0.7%
=========================================
Files 145 132 -13
Lines 6921 6143 -778
Branches 410 374 -36
=========================================
- Hits 5249 4616 -633
+ Misses 1672 1527 -145
Continue to review full report at Codecov.
|
| "CONFIG_whisk_timeLimit_min": "{{ limit_action_time_min | default() }}" | ||
| "CONFIG_whisk_timeLimit_max": "{{ limit_action_time_max | default() }}" | ||
| "CONFIG_whisk_timeLimit_std": "{{ limit_action_time_std | default() }}" | ||
| "CONFIG_whisk_timeLimit_timeoutfactor": "{{ limit_action_timeoutfactor | default() }}" |
Contributor
There was a problem hiding this comment.
Should this rather be part of the loadbalancer config since it is specific to the loadbalancer after all?
mhenke1
changed the title
markusthoemmes
added
the
review
If an active-ack does not appear after a certain timeout (action timeout + 1 minute), the active-ack is considered as lost and we "force" it to keep the loadbalancer's state sane. In some cases where the system gets increasingly slow, this timeout is too narrow. This makes the value configurable to be able to rapidly adjust it and to find a good default in production environments. Co-authored-by: Markus Thömmes <markusthoemmes@me.com>
Contributor
|
PG3 2561 🔵 |
BillZong
pushed a commit
to BillZong/openwhisk
that referenced
this pull request
Nov 18, 2019
If an active-ack does not appear after a certain timeout (action timeout + 1 minute), the active-ack is considered as lost and we "force" it to keep the loadbalancer's state sane. In some cases where the system gets increasingly slow, this timeout is too narrow. This makes the value configurable to be able to rapidly adjust it and to find a good default in production environments. Co-authored-by: Markus Thömmes <markusthoemmes@me.com>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
5 participants
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.
This PR introduces a configurable multiplicator to influence the forced active ack timeout period.
Description
In rare condition one can see that the load balancer put activations into queues of invokers falsely
identified to have free capacity because forced active acks are send out for activation that are waiting in the queue to be executed. With increasing the timeout the loadbalancer can be influenced to distribute the load wider in this cases.
Related issue and scope
My changes affect the following components
Types of changes
Checklist:
[PG3:2420 succeeded]