Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Disable downsampling for all aggregated namespaces if any namespace has downsample option is set to "all : false" #2

Merged
merged 184 commits into from
Feb 9, 2021

Conversation

anuprout
Copy link
Owner

@anuprout anuprout commented Feb 9, 2021

What this PR does / why we need it:

Fixes #

Special notes for your reviewer:

Does this PR introduce a user-facing and/or backwards incompatible change?:


Does this PR require updating code package or user-facing documentation?:


arnikola and others added 30 commits November 12, 2020 15:31
* Improe some slow tests

* lint

* lint

* goddamn imports
* Initial commit for proposed query handlers refactoring:
 * updated CustomHandler interface
 * unified how query routes are added. no need to pass router around and wrapping is done in one place.
 * routes are registered with names so they could be easily found later when custom handlers are registered.

* trying to make linter happy

* linter fixes

* revert old behaviour

* Make sure route methods are taken into account when adding and searching for named route.

* fixed code formatting

* [dbnode] Refactor wide query path (m3db#2826)

* [dbnode] Introduce Aggregator type (m3db#2840)

* [coordinator] Set default namespace tag to avoid colliding with commonly used "namespace" label (m3db#2878)

* [coordinator] Set default namespace tag to avoid colliding with common "namespace" default value

* Use defined constant

* Add downsampler test case to demonstrate override namespace tag

Co-authored-by: Wesley Kim <wesley@chronosphere.io>

* Improve some slow tests (m3db#2881)

* Improe some slow tests

* lint

* lint

* goddamn imports

* Changes after code review.

* [query] Remove dead code in prom package (m3db#2871)

* Register separate route for each method.

* linter fixes

* removed code duplication in hasndler_test

* Fail if route was already registered.

* formatted code

* Update src/query/api/v1/httpd/handler_test.go

Co-authored-by: Vilius Pranckaitis <vpranckaitis@gmail.com>

* Update src/query/api/v1/httpd/handler_test.go

Co-authored-by: Vilius Pranckaitis <vpranckaitis@gmail.com>

* More handler tests.

Co-authored-by: arnikola <artem@chronosphere.io>
Co-authored-by: Linas Medžiūnas <linasm@users.noreply.github.com>
Co-authored-by: Rob Skillington <rob.skillington@gmail.com>
Co-authored-by: Wesley Kim <wesley@chronosphere.io>
Co-authored-by: Vytenis Darulis <vytenis@uber.com>
Co-authored-by: Vilius Pranckaitis <vpranckaitis@gmail.com>
vpranckaitis and others added 26 commits January 11, 2021 09:39
* [coordinator] Rollout augmentM3Tags flag to true by default

* Fixup operator docs link

* Fixup doc site links
* refactored instrumentation code for peers bootstrapper.
added additional timers for peers bootstrapper to track individual steps.

* fixed linter warnings

* removed dead code

* removed dead code

* changes after review

* added new gauge - bootstrap-retries

* moved some instrumentation methods to instrumentationContexts for better encapsulation and easier usage.

* forgot to set nowFn

* added methods for creating instrumentation contexts.

* changes after code review
…ailure (m3db#3088)

* set bootstrap state to Bootstrapped when it returns no errors

* added bootstrap regression test.

* removed new test and updated old test to check for bootstrap states.
* Limit for time series read from disk

There is additional memory costs when time series need to be loaded from
disk (in addition to the actual bytes being read). This new limit allows
docLimits to be higher so series already in memory can easily be served
without hitting limits.
* [m3db] Check bloom filter before stream request allocation

* Add test assertions for bloom filer misses metric

* Remove redundant series-read metric
seekerMgr could change outside the lock, which might result in a nil
pointer
Check if there's at least one aggregated namespace with downsample disabled. If so, disabled the downsample object entirely
@anuprout anuprout merged commit 010a3e1 into anuprout:v1.0.0_Anup Feb 9, 2021
anuprout added a commit that referenced this pull request Feb 9, 2021
…espace has downsample option is set to "all : false" (#2)"

This reverts commit 010a3e1.
anuprout added a commit that referenced this pull request Feb 9, 2021
…espace has downsample option is set to "all : false" (#2)" (#3)

This reverts commit 010a3e1.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.