DOC-6872 Document StackExchange.Redis Active-Active geo-failover

[andy-stark-redis]

Description

Documents StackExchange.Redis's client-side geographic (Active-Active) failover support, shipped in the v3.1.0 GA release (StackExchange/StackExchange.Redis#3025).

Adds a new per-client page, dotnet/failover.md, modeled on the existing Jedis and redis-py failover pages, covering the StackExchange.Redis.Availability API:

• Connecting to a group of endpoints (ConnectGroupAsync), endpoint and group configuration
• Circuit breakers, automatic retries (WithRetry), and which operations are safe to retry
• Failback (IsConnected vs IsUnhealthy, FailbackDelay), manual failover (TryFailoverTo)
• Health checks (probe types, policies, per-member), managing members at runtime, monitoring, and Pub/Sub re-subscription

Also adds a StackExchange.Redis row to the supported-clients table on the failover overview page.

Notes for reviewers

• OSS Cluster failover cell — I marked this Yes in the overview table by inference (the multiplexer natively supports cluster endpoints, group members accept full ConfigurationOptions, and upstream's custom-probe docs handle cluster key routing), but upstream has no section stating it outright. Worth confirming with the SE.Redis maintainer before merge.
• GA, not preview — unlike the redis-py and Lettuce rows, 3.1.0 is a stable release, so this row has no "(Preview)" marker and the page carries no preview banner.
• Deliberately omitted the "Behavior when all endpoints are unhealthy" and "Troubleshooting" sections that Jedis/redis-py have — upstream docs don't provide sourced SE.Redis material for them yet.

🤖 Generated with Claude Code

---

Note

Low Risk
Documentation-only changes with no runtime, auth, or data-path impact.

Overview
Adds StackExchange.Redis client-side geographic failover documentation for the v3.1.0 StackExchange.Redis.Availability API, aligned with the existing Jedis/redis-py failover pages.

The new dotnet/failover.md page covers ConnectGroupAsync / ConnectionGroupMember groups, MultiGroupOptions (health checks, circuit breakers, retry policy), WithRetry() and command retry categories, failback (FailbackDelay, IsConnected / IsUnhealthy), runtime member management and TryFailoverTo(), ConnectionChanged monitoring, and Pub/Sub with automatic re-subscription. It also documents the SER007 / [Experimental] usage note for the availability namespace.

The failover overview (failover.md) links to the new page in relatedPages and adds a supported-clients table row listing basic failover, pub/sub failover, OSS cluster failover, and failback as Yes (GA, no preview marker).

^{Reviewed by Cursor Bugbot for commit ee2078b. Bugbot is set up for automated code reviews on this repo. Configure here.}

[github-actions]

🧠 Redis Memory

Found 5 related items from repository history:

• PR — DOC-6459 Integrate RS failover docs with client-specific stuff

  First step in integrating the general RS info about active-active health checks, failover, etc, with the info in the [clients section](https

• PR — DOC-6252 sections about failover behaviour when all endpoints are unhealthy

  Added info about this based on customer feedback. The corresponding section for the Lettuce geo failover page will be added in a separate PR

• PR — DOC-6810 Add Spring Data Redis client-side geo failover page

  A few things to check:

• PR — DOC-6689 added pub/sub failover details

• PR — DOC-6470 document StackExchange.Redis as the main C# client

  This basically reinterprets most of the existing NRedisStack docs as StackExchange.Redis and uses NRedisStack only for former Redis Stack fe

Memory updated at 533ca77

[github-actions]

Staging links:
https://redis.io/docs/staging/DOC-6872-dotnet-geo-failover/develop/clients/dotnet/failover/
https://redis.io/docs/staging/DOC-6872-dotnet-geo-failover/develop/clients/failover/

[github-actions]

DOC-6872

[github-actions]

🧠 Redis Memory

Found 5 related items from repository history:

• Commit — DOC-6872 Document StackExchange.Redis Active-Active geo-failover

  DOC-6872 Document StackExchange.Redis Active-Active geo-failover

• PR — DOC-6689 added pub/sub failover details

• PR — DOC-6459 Integrate RS failover docs with client-specific stuff

  First step in integrating the general RS info about active-active health checks, failover, etc, with the info in the [clients section](https

• PR — DOC-6252 sections about failover behaviour when all endpoints are unhealthy

  Added info about this based on customer feedback. The corresponding section for the Lettuce geo failover page will be added in a separate PR

• PR — DOC-6470 document StackExchange.Redis as the main C# client

  This basically reinterprets most of the existing NRedisStack docs as StackExchange.Redis and uses NRedisStack only for former Redis Stack fe

Memory updated at 58af957

[dwdougherty]

Language LGTM.

[andy-stark-redis]

Thanks @dwdougherty !

[mgravell]

unlike the redis-py and Lettuce rows, 3.1.0 is a stable release,

The release is stable / GA, but we've left a small [Experimental("SER007")], which links here - I don't know if this changes things at all.

Ignore me, I see you've covered that.

[mgravell]

I guess there's always going to be a lot of duplication between these docs and the library docs, but: there's a lot of duplication here; a small part of me wonders whether "keep minimal, and link" is better, but: obviously please follow whatever the normal approach is

[github-actions]

🧠 Redis Memory

Found 5 related items from repository history:

• Commit — DOC-6872 Document StackExchange.Redis Active-Active geo-failover

  DOC-6872 Document StackExchange.Redis Active-Active geo-failover

• PR — DOC-6689 added pub/sub failover details

• PR — DOC-6459 Integrate RS failover docs with client-specific stuff

  First step in integrating the general RS info about active-active health checks, failover, etc, with the info in the [clients section](https

• PR — DOC-6470 document StackExchange.Redis as the main C# client

  This basically reinterprets most of the existing NRedisStack docs as StackExchange.Redis and uses NRedisStack only for former Redis Stack fe

• PR — DOC-6252 sections about failover behaviour when all endpoints are unhealthy

  Added info about this based on customer feedback. The corresponding section for the Lettuce geo failover page will be added in a separate PR

Memory updated at ee2078b

[andy-stark-redis]

I guess there's always going to be a lot of duplication between these docs and the library docs, but: there's a lot of duplication here; a small part of me wonders whether "keep minimal, and link" is better, but: obviously please follow whatever the normal approach is

Yeah, this has always been an issue with some of these client-specific docs. I'm currently implementing some changes in the docs build so that our source files can be more like "standard" Markdown (they currently have loads of Hugo-specific stuff in them). This should hopefully give us the option of keeping all the client-specific docs in the SE.Redis repo and sharing the content between redis.io and the SE.Redis library docs. Note that I'm not pushing for this approach or recommending it yet but it might be something we could consider in the future if it's useful.

Anyway, for now I'll continue with the existing approach for the geo failover docs, but thanks for sharing your thoughts about this.