GitHub Repo
Open WebUI Community Reports Sentinel Name Resolution Bug That May Cause Redis Connection Failures
A community report from October 4 involving Open WebUI 0.11.4 says Redis Sentinel master group names may be lowercased, causing service lookup failures. Official documentation and Python parsing rules support this analysis, but the upstream fix status remains unconfirmed.

On October 4, the Open WebUI community reported that Redis connections may fail when using Redis Sentinel in a Kubernetes deployment with version 0.11.4 if the master group name contains uppercase letters. The reporter’s example was myMaster: Sentinel can find this group, but Open WebUI searches for mymaster and raises MasterNotFoundError. This points to a possible troubleshooting path, but so far it is a community reproduction and has not been confirmed by maintainers. Issue report
The key is that the same URL field carries different meanings. The official Open WebUI documentation specifies that when REDIS_SENTINEL_HOSTS is configured, the hostname in REDIS_URL is interpreted as the Sentinel service name, rather than a regular network hostname. As a result, engineers may follow the documentation and enter the group name but still encounter a mismatch between the parsed string and the Sentinel configuration. Environment variables documentation
The report points to the use of parts.hostname in parse_redis_url() as the cause. Python’s official documentation explicitly states that the hostname in a URL parsing result is converted to lowercase. This is standard handling for hostnames, but it may alter a service identifier whose original case should be preserved. This case shows that when applying a URL parser to a logical name, engineers need to check whether normalization changes its application-level meaning. Python parsing documentation
The technical impact mainly concerns deployments that require shared state. Open WebUI’s scaling documentation describes using Redis to coordinate multiple processes or instances and provides a Sentinel high-availability configuration. A service name resolution failure could therefore interfere with these architectures. This is a deployment risk inferred from the documentation; the available report does not establish that all multi-instance environments are affected. Scaling documentation
Engineers can first compare the group name in the error log with the actual Sentinel name, and include capitalization in connection testing. Next, they should watch for whether upstream preserves the service name’s original case, adds a mixed-case regression test, and includes a fix in an official release. At the time of review, the issue remained open and the page listed no related fix PR; renaming should not be treated as a solution that can be applied directly in every environment. Issue status