REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines #44

scott-parillo · 2025-12-06T02:13:39Z

Summary

The environment size was updated and now defines X-Large. Collapsed multiple tables into a single Elastic Stack infrastructure table to describe the infrastructure recommendations using the Server 2025 GA EW production certification results.
Retension Policy Guideline page added.
Environment Watch Performance Impact page added.

Test Evidence for Retention Policy Guidelines - SQE: #44 (review)
Test Evidence for Retention Policy Guidelines - DEV: 12_16_2025_Retention_Guideline_Dev_Evidence.docx

…. Collapsed multiple tables into a single Elastic Stack infrastructure table to describe the infrastructure recommendations using the Server 2025 GA EW production certification results.

dinesh1010101 · 2025-12-08T15:02:04Z

docs/elasticsearch_pre_installation_overview.md

-| Kibana                                    | 1                      | 4          |
-| APM Server                                | 1                      | 4          |
-
+| Environment Size              | Web Servers | Agent Servers | Worker Servers | SQL Distributed Servers |


@scott-parillo This is mismatching with the slack thread mentioned numbers(https://kcura-pd.slack.com/archives/C0616SVFYBU/p1764962920595329?thread_ts=1764959199.337929&cid=C0616SVFYBU)

Also mismatching with our initial table here

May i know whether these numbers are intentional or do i have make some changes?

dinesh1010101 · 2025-12-10T11:45:12Z

docs/elasticsearch_pre_installation_overview.md

 **A few other key notes and reminders:**

 - **Tuning for speed** – Review Elastic’s guidance on how to tune the environment for speed [here](https://www.elastic.co/guide/en/elasticsearch/reference/current/tune-for-search-speed.html).
- **Hosting Elastic** – While the guidance below recommends installing the Elastic components on many dedicated servers, there are no hard requirements to isolate Elasticsearch, Kibana, or APM Server on dedicated hosts. As evident with the Development environment specifications, the full Elastic stack can be deployed on a single host if that server can meet the storage needs.


Removed as per the feedback: https://kcura-pd.slack.com/archives/C0616SVFYBU/p1764987926224259?thread_ts=1764987546.162489&cid=C0616SVFYBU

…olume-reduction

dinesh1010101 · 2025-12-15T07:13:06Z

docs/environment_watch_performance_impact.md

+| **Processing** | **+450% faster** | Processing performance has improved dramatically, delivering a 450% speed increase that will noticeably accelerate end-to-end workflows. |
+| **Review (Conversion)** | **+5% faster** | Review operations saw a modest 5% improvement, providing slightly faster document conversion without any workflow disruption. |
+| **Imaging & Production** | **Stable (±4%)** | Imaging and production performance remained stable, with changes within a ±4% range, resulting in no meaningful impact to customer workflows. |
+| **Data Transfer** | **Mixed results** | Native file operations improved by 4–38%, offering smoother import/export performance. Image-based workflows saw some declines—most notably a 157% slowdown in RIP image export—which may impact image-heavy projects. |


Note: Here we are disclosing 157% slowdown which might give negative perspective to client

I believe that huge performance increases/decreases have yet to be vetted by the teams. Honestly, it defies logic but I don't want to completely discount. Until this has been thoroughly investigated and a clear conclusion drawn, I would advise removing any such results (positive or negative) until then. As it's currently worded, this sounds concerning but the results indicate otherwise.

Removed the info as of now until other team comes to a conclusion
966d848

dinesh1010101 · 2025-12-15T07:14:30Z

docs/environment_watch_performance_impact.md

+
+## Conclusion
+
+Environment Watch delivers significant performance improvements for processing workloads while maintaining stable performance for most other Relativity operations. Organizations with heavy image-based data transfer workflows should evaluate their specific use cases to ensure alignment with their performance requirements.


Note: Depends on decision in including RIP image slowdown info, last line need to be modified/removed.

Removed the info as of now until other team comes to a conclusion
966d848

Rahiman-Nadaf

Reviewed retention policy guidelines and looks good to me

docs/environment-watch/post-install-verification/retention-policy.md

docs/elasticsearch_pre_installation_overview.md

docs/environment-watch/post-install-verification/retention-policy.md

docs/environment-watch/post-install-verification.md

docs/elasticsearch_pre_installation_overview.md

docs/elasticsearch_retention_policy_guidelines.md

DaRealRahul1

I have completed reviewing the document. Could you please check the document below and make the necessary changes accordingly.
https://jira.kcura.com/secure/attachment/758132/EW%20Review.docx

dinesh1010101 · 2025-12-17T06:47:56Z

Thanks @DaRealRahul1 , i have accommodated all your feedback, Please find the list below and respective action.

Mention the steps exactly where need to navigate for Kibana Dev Console - Added steps in 1st step, next steps haven't repeat the same as it will understood in the 1st step itself.
The output is different I can see the extra entry "mode": "standard" in the above doc but in my TestVm I cant see it, as well as the entries present in "composed_of ": is in not in series, If it is expected please mention the same - As it mentioned sample, its already covering it.
Better to mention the GET and PUT command for metrics-apm.app@template and traces-apm@template - Corrected as i also felt the same from user perspective.
In above step its not mention to update the highlighted template according to the template need to be update for metrics-apm.app@template and traces-apm@template - Invalid as i have added seperate steps for metrics and traces as well.
No need to update as the lifecycle is already 90 - Content updated.
Same for this as well mention the step like step 1 - Datastream - Mentioned in initial stage on steps to navigate to DEV console
The outputs are different for all the below commands compared to the above document, please correct it. - Corrected
Where we can check data lifecycle management is active, mention the steps for this as well if possible - Altered wordings.

Requesting you to review and approve the PR.

docs/environment-watch/post-install-verification/retention-policy.md

docs/elasticsearch_retention_policy_guidelines.md

…ormance impact

dinesh1010101 · 2025-12-19T05:00:11Z

As discussed with @KarunaDhawan , removed the performance impact info. Until teams conclude that information, this PR will be blocked.

CC: @Phani61284 @anujavinash @satishdayala-relativity

KarunaDhawan · 2026-01-14T20:07:27Z

docs/elasticsearch_retention_policy_guidelines.md

+
+These guidelines define retention policies for logs, metrics, and traces collected in Elasticsearch and viewed through Kibana. Proper retention management is critical for:
+
+- **Storage Optimization** – Prevents excessive disk usage by automatically removing outdated data


Storage and cost are tied together, will be best to combine them here:
Storage Optimization & Cost Control – Prevents excessive disk usage and reduces infrastructure costs by automatically removing outdated data

KarunaDhawan · 2026-01-14T20:20:52Z

docs/elasticsearch_retention_policy_guidelines.md

+```
+Docs/Day (Daily Documents) = 6M + (Web_Server_Count × 2M) + (Agent_Server_Count × 2M) + (Worker_Server_Count × 400k) + (SQL_Distributed_Server_Count × 500k)
+
+GiB/Day (Daily Storage) = Docs/Day × 380 / 1024³


where does this 380 comes from? Perhaps we should explain what that number is?

Sure, updated the same.

KarunaDhawan · 2026-01-14T20:53:54Z

docs/elasticsearch_retention_policy_guidelines.md

+### Step 3: Delete Existing Data Streams (Setup Time Only)
+
+> [!WARNING]
+> This step should only be performed once during initial setup. Deleting data streams will permanently remove all data and indices under those data streams.


The data stream deletion step is destructive. Can we make the warning more prominent or clarify when it is safe to run?

Sure, Updated the same.

KarunaDhawan · 2026-01-14T21:02:49Z

docs/elasticsearch_retention_policy_guidelines.md

+**Sample Request:**
+
+```
+# Here logs-apm.app@template is the name of the index template


The Logs, Metrics, and Traces sections repeat the same workflow. Can we describe the pattern once and only show full JSON once?

Initially, I had kept it the same way. However, during SQE verification, I received feedback that they felt stuck or confused and requested that all three be shown separately, so users can follow more smoothly. However, please let me know if you still prefer combining them into a single section.

@dinesh1010101 If compressing into one JSON causes confusion, perhaps we can keep it how it is set up currently, thanks

KarunaDhawan · 2026-01-26T20:05:07Z

docs/elasticsearch_retention_policy_guidelines.md

+### Purpose
+
+These guidelines define retention policies for logs, metrics, and traces collected in Elasticsearch and viewed through Kibana. Proper retention management is critical for:
+


@dinesh1010101 Please add EW context clarifying that configuring Elasticsearch retention policies is optional. Environment Watch operates correctly using default retention settings, and this configuration should be applied only when customers need to customize retention behavior. This helps prevent readers from assuming the configuration is required.
Note text could be something like this: Configuring Elasticsearch retention policies is optional. Environment Watch works out of the box using default retention settings. The configurations described here should be applied only if you need to customize how long data is retained to align with your organization’s storage, performance, or compliance requirements.

KarunaDhawan · 2026-01-26T20:12:05Z

docs/elasticsearch_retention_policy_guidelines.md

+**Sample Request:**
+
+```
+# Here logs-apm.app@template is the name of the index template


@dinesh1010101 If compressing into one JSON causes confusion, perhaps we can keep it how it is set up currently, thanks

KarunaDhawan · 2026-01-26T20:16:27Z

docs/elasticsearch_retention_policy_guidelines.md

+> - You are performing **initial setup** and no production data exists yet
+> 
+> **Do NOT run this on production systems with active data.**
+


Thanks for adding the Note but still it seems like this is a necessary step by the users. Pls update this to explicitly mark the step as optional and intended only for initial setup or controlled scenarios. For example:

Step 3: Delete Existing Data Streams (Setup Time Only)

⚠️ DESTRUCTIVE OPERATION – PERMANENT DATA LOSS

This step is optional and is not required for most Environment Watch deployments. It should only be performed during initial setup or in controlled, non-production scenarios.

This step will permanently delete all data and indices in the specified data streams. There is no recovery. Only proceed if:

You are in a development or non-production environment, OR

You have backed up all critical data from these data streams, OR

You are performing initial setup and no production data exists yet

Do NOT run this on production systems with active data.

KarunaDhawan · 2026-01-26T20:21:32Z

docs/elasticsearch_retention_policy_guidelines.md

+}
+```
+
+> [!IMPORTANT]


This note is great but will be helpful to have it right before the Index template update steps.

KarunaDhawan · 2026-01-26T20:28:03Z

docs/environment_watch_performance_impact.md

+
+## Overview
+
+This document provides transparent information about the performance overhead Environment Watch introduces to standard Relativity workloads, based on comprehensive testing in a production-like environment.


Pls update this to: This document provides transparent information about the performance overhead Environment Watch introduces to standard Relativity workloads, based on testing conducted in a controlled, production-like environment. Actual performance may vary depending on workload characteristics, environment size, infrastructure configuration, and usage patterns.

KarunaDhawan · 2026-01-26T20:28:50Z

docs/environment_watch_performance_impact.md

+| **Review (Conversion)** | **+5% faster** | Review operations saw a modest 5% improvement, providing slightly faster document conversion without any workflow disruption. |
+| **Imaging & Production** | **Stable (±4%)** | Imaging and production performance remained stable, with changes within a ±4% range, resulting in no meaningful impact to customer workflows. |
+| **Data Transfer** | **~5-6% faster on average** | Data transfer operations showed performance improvements with Imports demonstrating ~10% faster performance on average, while Exports (excluding RIP Images Export) were ~1% faster on average, resulting in an approximate 5-6% overall improvement. |
+


Pls add a note before the table: The results below reflect observed outcomes from internal testing and are provided for transparency. These results should not be interpreted as guaranteed performance improvements for all Environment Watch deployments.

KarunaDhawan · 2026-01-26T20:31:11Z

docs/environment_watch_performance_impact.md

+
+## Conclusion
+
+Environment Watch has demonstrated minimal to positive impact on Relativity workloads across comprehensive testing. Most operations showed performance improvements, with Processing, Data Transfer, and Review all performing faster. Imaging and Production workflows remained stable. These results confirm that Environment Watch provides valuable observability and monitoring capabilities without compromising your Relativity system's performance.


Pls update this to something like this as the results may vary in diff env: Environment Watch has demonstrated minimal to positive impact on Relativity workloads based on comprehensive testing in a controlled, production-like environment. Most operations showed performance improvements, with Processing, Data Transfer, and Review performing faster, while Imaging and Production workflows remained stable. Environment Watch is designed to deliver observability and monitoring capabilities with minimal overhead; however, actual performance results may vary based on customer-specific configurations, environment size, and workload characteristics.

REL-1207540: The environment size was updated and now defines X-Large…

341f309

…. Collapsed multiple tables into a single Elastic Stack infrastructure table to describe the infrastructure recommendations using the Server 2025 GA EW production certification results.

dinesh1010101 reviewed Dec 8, 2025

View reviewed changes

dinesh1010101 added 2 commits December 10, 2025 17:09

REL-1224050: Basic Retension Content

62efbac

REL-1207540: Feedback fix

e70d829

dinesh1010101 reviewed Dec 10, 2025

View reviewed changes

dinesh1010101 added 3 commits December 12, 2025 19:25

REL-1224050: More section added

7323447

REL-1207540: performance impact md file added

cd4b49d

Merge branch 'REL-1224050-Retension-SOP' into REL-1207540-telemetry-v…

93329dd

…olume-reduction

dinesh1010101 mentioned this pull request Dec 15, 2025

REL-1224050: Retension Policy SOP (WIP) #47

Closed

dinesh1010101 changed the title ~~REL-1207540-telemetry-volume-reduction~~ REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines Dec 15, 2025

dinesh1010101 reviewed Dec 15, 2025

View reviewed changes

dinesh1010101 added 4 commits December 15, 2025 15:34

REL-1207540: Self feedback/observation fix

b1908d6

REL-1224050: Banner added in rentention policy verification

f5f4f81

REL-1224050: Spelling fix

0434011

REL-1207540: Feedback fix in Licensing

ae1187c

dinesh1010101 marked this pull request as ready for review December 15, 2025 10:28

dinesh1010101 requested a review from a team as a code owner December 15, 2025 10:28

dinesh1010101 requested review from KarunaDhawan, Rahiman-Nadaf, anujavinash and satishdayala-relativity December 15, 2025 10:29

Rahiman-Nadaf approved these changes Dec 15, 2025

View reviewed changes

dinesh1010101 requested a review from DaRealRahul1 December 16, 2025 05:56

dinesh1010101 added 3 commits December 16, 2025 14:28

REL-1224050: Delete existing stream step added

aba3989

REL-1224050: Red color highlight removal

f6b2a64

REL-1224050: Red color highlight removal

8a36ce1

anujavinash requested a review from dinesh1010101 December 16, 2025 12:36

anujavinash requested changes Dec 16, 2025

View reviewed changes

dinesh1010101 added 3 commits December 16, 2025 20:15

REL-1224050: Minor feedback fix

aea42c9

REL-1224050: Feedback fix

2d81e28

REL-1224050: Feedback fix

ccf2acf

anujavinash approved these changes Dec 16, 2025

View reviewed changes

DaRealRahul1 suggested changes Dec 16, 2025

View reviewed changes

REL-1224050: Feedback fix from SQE

fb066b5

REL-1224050: Feedback fix from SQE

efba91f

DaRealRahul1 approved these changes Dec 17, 2025

View reviewed changes

docs/environment-watch/post-install-verification/retention-policy.md Show resolved Hide resolved

docs/elasticsearch_retention_policy_guidelines.md Show resolved Hide resolved

DaRealRahul1 approved these changes Dec 17, 2025

View reviewed changes

satishdayala-relativity approved these changes Dec 17, 2025

View reviewed changes

dinesh1010101 removed their request for review December 18, 2025 04:21

REL-1207540: TDB placeholder added for summary and conclusion in perf…

966d848

…ormance impact

dinesh1010101 changed the title ~~REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines~~ (DO NOT MERGE) - REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines Dec 19, 2025

anujavinash changed the base branch from main to REL-1238635-cumulative-cum-folder-hierarchy-change December 22, 2025 10:33

KarunaDhawan requested changes Jan 14, 2026

View reviewed changes

dinesh1010101 added 2 commits January 22, 2026 15:21

REL-1207540: TBD filled with latest info in EW Performance impact page

6e52820

REL-1207540: PM feedback fix

3beef6a

dinesh1010101 changed the title ~~(DO NOT MERGE) - REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines~~ REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines Jan 23, 2026

KarunaDhawan requested changes Jan 26, 2026

View reviewed changes


		## Conclusion

		Environment Watch delivers significant performance improvements for processing workloads while maintaining stable performance for most other Relativity operations. Organizations with heavy image-based data transfer workflows should evaluate their specific use cases to ensure alignment with their performance requirements.


		These guidelines define retention policies for logs, metrics, and traces collected in Elasticsearch and viewed through Kibana. Proper retention management is critical for:

		- Storage Optimization – Prevents excessive disk usage by automatically removing outdated data

		### Purpose

		These guidelines define retention policies for logs, metrics, and traces collected in Elasticsearch and viewed through Kibana. Proper retention management is critical for:


		## Overview

		This document provides transparent information about the performance overhead Environment Watch introduces to standard Relativity workloads, based on comprehensive testing in a production-like environment.


		## Conclusion

		Environment Watch has demonstrated minimal to positive impact on Relativity workloads across comprehensive testing. Most operations showed performance improvements, with Processing, Data Transfer, and Review all performing faster. Imaging and Production workflows remained stable. These results confirm that Environment Watch provides valuable observability and monitoring capabilities without compromising your Relativity system's performance.

REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines #44

Are you sure you want to change the base?

REL-1207540-telemetry-volume-reduction & REL-1224050-Retension-Policy-Guidelines #44

Uh oh!

Conversation

scott-parillo commented Dec 6, 2025 • edited by dinesh1010101 Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Summary

Uh oh!

dinesh1010101 Dec 8, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Rahiman-Nadaf left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

DaRealRahul1 left a comment

Choose a reason for hiding this comment

Uh oh!

dinesh1010101 commented Dec 17, 2025

Uh oh!

Uh oh!

Uh oh!

dinesh1010101 commented Dec 19, 2025

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

scott-parillo commented Dec 6, 2025 •

edited by dinesh1010101

Loading

dinesh1010101 Dec 8, 2025 •

edited

Loading