jtuglu1 opened a new pull request, #19303:
URL: https://github.com/apache/druid/pull/19303
### Description
<!-- Describe the goal of this PR, what problem are you fixing. If there is
a corresponding issue (referenced above), it's not necessary to repeat the
description here, however, you may choose to keep one summary sentence. -->
<!-- Describe your patch: what did you change in code? How did you fix the
problem? -->
<!-- If there are several relatively logically separate changes in this PR,
create a mini-section for each of them. For example: -->
We've seen issues with the `cost` segment balancing strategy not working
correctly in production, where it leaves historicals in a tier in a bimodal
distribution: ~50% in ~100% utilization and ~50% in 60% utilization. This is
due to the cost-based balancer not having any visibility into disk util on each
historical. The diskNormalized strategy is available, but is marked as not
"production-ready," I believe due to the following reason:
The “diskNormalized” balancer works by extending from the “cost” balancer
and overriding the computeCost function in the following way:
```
double cost = super.computeCost(proposalSegment, server,
includeCurrentServer);
if (cost == Double.POSITIVE_INFINITY) {
return cost;
}
int nSegments = 1;
if (server.getServer().getNumSegments() > 0) {
nSegments = server.getServer().getNumSegments();
}
double normalizedCost = cost / nSegments;
double usageRatio = (double) server.getSizeUsed() / (double)
server.getServer().getMaxSize();
return normalizedCost * usageRatio;
```
This logic is a bit strange. Consider the case where historicals are all
homogeneous, so server.getServer().getMaxSize() can be treated as a constant.
In that case, the returned value simplifies to:
```
normalizedCost * usageRatio
= normalizedCost * usedSize / maxSize
= normalizedCost * usedSize / CONST
= cost / numSegments * usedSize / CONST
= cost * (usedSize / numSegments) / CONST
= cost * avgSegmentSize / CONST
```
So, in practice, it’s just adjusting cost by a factor of avgSegmentSize.
This could lead to imbalances that make themselves worse over time, especially
if newly created segments tend to have smaller sizes than older ones (which is
common with append-based ingestion methods).
<!--
In each section, please describe design decisions made, including:
- Choice of algorithms
- Behavioral aspects. What configuration values are acceptable? How are
corner cases and error conditions handled, such as when there are insufficient
resources?
- Class organization and design (how the logic is split between classes,
inheritance, composition, design patterns)
- Method organization and design (how the logic is split between methods,
parameters and return types)
- Naming (class, method, API, configuration, HTTP endpoint, names of
emitted metrics)
-->
<!-- It's good to describe an alternative design (or mention an alternative
name) for every design (or naming) decision point and compare the alternatives
with the designs that you've implemented (or the names you've chosen) to
highlight the advantages of the chosen designs and names. -->
<!-- If there was a discussion of the design of the feature implemented in
this PR elsewhere (e. g. a "Proposal" issue, any other issue, or a thread in
the development mailing list), link to that discussion from this PR description
and explain what have changed in your final design compared to your original
proposal or the consensus version in the end of the discussion. If something
hasn't changed since the original discussion, you can omit a detailed
discussion of those aspects of the design here, perhaps apart from brief
mentioning for the sake of readability of this PR description. -->
<!-- Some of the aspects mentioned above may be omitted for simple and small
changes. -->
#### Release note
<!-- Give your best effort to summarize your changes in a couple of
sentences aimed toward Druid users.
If your change doesn't have end user impact, you can skip this section.
For tips about how to write a good release note, see [Release
notes](https://github.com/apache/druid/blob/master/CONTRIBUTING.md#release-notes).
-->
Fix diskNormalized cost balancer strategy
<hr>
<!-- Check the items by putting "x" in the brackets for the done things. Not
all of these items apply to every PR. Remove the items which are not done or
not relevant to the PR. None of the items from the checklist below are strictly
necessary, but it would be very helpful if you at least self-review the PR. -->
This PR has:
- [ ] been self-reviewed.
- [ ] using the [concurrency
checklist](https://github.com/apache/druid/blob/master/dev/code-review/concurrency.md)
(Remove this item if the PR doesn't have any relation to concurrency.)
- [ ] added documentation for new or modified features or behaviors.
- [ ] a release note entry in the PR description.
- [ ] added Javadocs for most classes and all non-trivial methods. Linked
related entities via Javadoc links.
- [ ] added or updated version, license, or notice information in
[licenses.yaml](https://github.com/apache/druid/blob/master/dev/license.md)
- [ ] added comments explaining the "why" and the intent of the code
wherever would not be obvious for an unfamiliar reader.
- [ ] added unit tests or modified existing tests to cover new code paths,
ensuring the threshold for [code
coverage](https://github.com/apache/druid/blob/master/dev/code-review/code-coverage.md)
is met.
- [ ] added integration tests.
- [ ] been tested in a test Druid cluster.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
