alberttwong commented on code in PR #11919: URL: https://github.com/apache/nifi/pull/11919#discussion_r4617159126
########## nifi-extension-bundles/nifi-redis-bundle/nifi-redis-extensions/src/test/java/org/apache/nifi/redis/service/ITRedisClusterDistributedMapCacheClientService.java: ########## @@ -0,0 +1,29 @@ +/* + * Licensed to the Apache Software Foundation (ASF) under one or more + * contributor license agreements. See the NOTICE file distributed with + * this work for additional information regarding copyright ownership. + * The ASF licenses this file to You under the Apache License, Version 2.0 + * (the "License"); you may not use this file except in compliance with + * the License. You may obtain a copy of the License at + * + * http://www.apache.org/licenses/LICENSE-2.0 + * + * Unless required by applicable law or agreed to in writing, software + * distributed under the License is distributed on an "AS IS" BASIS, + * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + * See the License for the specific language governing permissions and + * limitations under the License. + */ +package org.apache.nifi.redis.service; + +/** + * Runs the atomic cache contract against the Redis Cluster API, with all hash slots assigned to one node. + * This exercises cluster command routing without requiring fixed Docker host ports or a multi-node network. + */ +public class ITRedisClusterDistributedMapCacheClientService extends ITRedisDistributedMapCacheClientService { Review Comment: You're right: the committed one-node Cluster API fixture does not exercise `MOVED` or `ASK`. That gap is now explicit in the PR description; the Javadoc clarification remains a source follow-up. There are two separate routing paths. In **Cluster mode / OSS Cluster API**, the client discovers the hash-slot topology, connects to the relevant node proxies, and handles redirects when necessary. In **Standalone mode**, NiFi connects to a database proxy endpoint and Redis routes requests across backend shards and nodes. Standalone mode therefore supports sharded, multi-node Redis Enterprise Software and Redis Cloud deployments. Cluster mode is an optional routing optimization, not a requirement for that deployment. Official references: [proxy routing](https://redis.io/docs/latest/operate/rs/databases/configure/proxy-policy/), [OSS Cluster API architecture](https://redis.io/docs/latest/operate/rs/clusters/optimize/oss-cluster-api/), and [Cloud clustering](https://redis.io/docs/latest/operate/rc/databases/configuration/clustering/). I have now separately validated the proxy path against Software with 1 node / 1 shard, 1 node / 3 shards, 3 nodes / 1 replicated primary, and 3 nodes / 3 primaries, plus Cloud Pro with 1 and 2 primary shards and multi-zone replication. Software planned failover, abrupt primary-node loss, and continuous atomic operations during one-to-three-shard expansion also passed through the proxy endpoint. The new external-endpoint harnesses are local acceptance artifacts, not committed tests. These results establish proxy routing and recovery coverage; they do not establish multi-node Cluster API redirect handling or global Active-Active compare-and-set. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
