9.1 Test Plan

[rubvs]

Manual Test Plan

When picking up a test case, please add your name to this overview beforehand and tick the checkbox when finished.
Testing can be started when the first build candidate (BC) is available in the CFT region.

Smoke Testing ESS setup

Thanks to #8303 further smoke tests are run automatically on ESS now.
Consider extending the smoke tests to include more test cases which we'd like to cover

ES apm-data plugin

• apm-data: Use representative count as event.success_count if available elasticsearch#119995 @isaacaflores2
• feat: enable date_detection for all apm data streams elasticsearch#128913 @1pkg @rubvs
• [apm-data] Set event.dataset if empty for logs elasticsearch#129074 @rubvs
• Update traces duration mappings with appropriate unit type elasticsearch#129418 @1pkg

apm-aggregation

List of changes: elastic/apm-aggregation@v1.2.0...v1.3.0

• Optimize range tombstones creation apm-aggregation#258 @rubvs

go-docappender

List of changes: elastic/go-docappender@v2.4.0...v2.10.0

• feat: remove deprecated tracer config go-docappender#247 @rubvs
• Failure Store: Update go-docappender to respect failure store status go-docappender#228 @rubvs

apm-data

List of changes: elastic/apm-data@v1.16.0...v1.19.2

• otlp: handle host.ip attribute data type more defensively apm-data#500 @rubvs
• feat: add readBatch span apm-data#492 @rubvs
• feat: replace go agent tracer with otel apm-data#487 @rubvs
• fix: Truncate string slice attributes in OTLP labels apm-data#434 @rubvs

apm-server

List of changes: v9.0.0...main

• feat: use go-freelru consistently #16549 @1pkg
• Change server information endpoint / to only accept GET and HEAD requests #15976 @rubvs
• apm: Add config for tail-based sampling discard on write integrations#13950 @rubvs
• [APM] integration new tail sampling settings kibana#224479 @rubvs
• Add sampling.tail.ttl to sample yml #16579 @isaacaflores2
• fix: add cacerts to ubi docker image #16928 @1pkg

[rubvs]

Tested

• Tested Change server information endpoint / to only accept GET and HEAD requests #15976 in Change server information endpoint / to only accept GET and HEAD requests #15976 (comment)
• Tested fix: Truncate string slice attributes in OTLP labels apm-data#434 in fix: Truncate string slice attributes in OTLP labels apm-data#434 (comment)
• Tested feat: replace go agent tracer with otel apm-data#487 & feat: add readBatch span apm-data#492 in feat: replace go agent tracer with otel apm-data#487 (comment)
• Tested otlp: handle host.ip attribute data type more defensively apm-data#500 in otlp: handle host.ip attribute data type more defensively apm-data#500 (comment)
• Tested feat: remove deprecated tracer config go-docappender#247 in feat: remove deprecated tracer config go-docappender#247 (comment)
• Tested Failure Store: Update go-docappender to respect failure store status go-docappender#228 when having implemented https://github.com/elastic/apm-managed-service/pull/1614
• Tested apm: Add config for tail-based sampling discard on write integrations#13950 with apm: Add config for tail-based sampling discard on write integrations#13950 (comment)
• Tested [APM] integration new tail sampling settings kibana#224479 with [APM] integration new tail sampling settings kibana#224479 (comment)
• Tested [apm-data] Set event.dataset if empty for logs elasticsearch#129074 with [apm-data] Set event.dataset if empty for logs elasticsearch#129074 (comment)

Removed

• tbs: Update storage metrics to be reported synchronously in the existing runDiskUsageLoop method #17154, since monitoring: TBS lsm_size and value_log_size are reported less frequently than other TBS metrics #17104 is still ongoing.
• fix: do not overwrite registries when publishing otel metrics to beats #15517, already added to 8.x and the unit test suffices.
• Add BulkIndexerPool to improve resource sharing go-docappender#243 confirmed with @marclop & feat: remove reslice logic apm-data#497 confirmed with @axw
• Confirmed with @carsonip to remove Add config PopulateFailedDocsSource to bulk indexer go-docappender#236
• Confirmed with @kruskall to remove fix: do not copy binary on explicit task and fix race condition #16077
• Allow passing a custom item action go-docappender#231, since it's not used by APM Server, see Allow passing a custom item action go-docappender#231 (comment)
• Add BulkIndexerItem.RequireDataStream go-docappender#235, since it's not used by APM Server
• feat: add include_source_on_error param to bulk index requests go-docappender#229, since it's already in 9.0.2
• Make IndexTemplateRegistry project-aware elasticsearch#126986 not used in APM Server
• Make use of new projectClient method and remove old one elasticsearch#129393 not used in APM Server

Shipped in 9.0.0

• TBS: set default sampling.tail.storage_limit to 0 but limit disk usage to 90% #15467
• TBS: make storage_limit follow processor lifecycle; update TBS processor config #15488
• TBS: change default disk usage threshold to 0.8 #15524

Shipped in 9.0.3

• TBS: Optimize performance for instances with more CPU and memory #17254
• TBS: Fix incorrectly large pebble database lsm_size monitoring metric #17149

[rubvs]

To what versions are we backporting elastic/apm-data#434?

[rubvs]

Confirmed with @lahsivjar that we want to test elastic/apm-aggregation#258

[isaacaflores2]

I will test #17154
We will still need to test #17154

• Changes for 9.X branches have been merged
• 8.19 work is ongoing

[isaacaflores2]

Marked #16579 as complete since it was only changes to sample yaml files. Changes are present in the main branch

[rubvs]

There is an issue with elastic/elasticsearch#128913. I left a comment here elastic/elasticsearch#128913 (comment)

[rubvs]

Removed elastic/apm-aggregation#269 based on the weekly convo.

[rubvs]

Left a comment in elastic/apm-aggregation#258 (comment) for elastic/apm-aggregation#258. Asking @lahsivjar to double check my writeup.