suvodeep-pyne commented on code in PR #17099: URL: https://github.com/apache/pinot/pull/17099#discussion_r2487998175
########## pinot-segment-local/src/main/java/org/apache/pinot/segment/local/utils/ServerReloadJobStatusCache.java: ########## @@ -0,0 +1,99 @@ +/** + * 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.pinot.segment.local.utils; + +import com.google.common.cache.Cache; +import com.google.common.cache.CacheBuilder; +import java.util.concurrent.TimeUnit; +import javax.annotation.Nullable; +import javax.annotation.concurrent.ThreadSafe; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import static java.util.Objects.requireNonNull; + + +/** + * In-memory cache for tracking reload job status on server side. + * Phase 1: Only tracks failure count per job. + * + * <p>Thread-safe for concurrent access. Uses Guava Cache with LRU eviction + * and time-based expiration. + */ +@ThreadSafe +public class ServerReloadJobStatusCache { + private static final Logger LOG = LoggerFactory.getLogger(ServerReloadJobStatusCache.class); + + private final Cache<String, ReloadJobStatus> _cache; + + /** + * Creates a cache with the given configuration. + * + */ + public ServerReloadJobStatusCache() { + final ServerReloadJobStatusCacheConfig config = new ServerReloadJobStatusCacheConfig(); Review Comment: yes. I think having some knobs like size of the reload cache in terms of # of jobs, or history length in days etc are helpful to keep the cache in check. or may be simply turning this off! hardcoding these params didn't seem the right approach. so, adding cache configurability will be part of the next set of upcoming changes. > do ensure this is well tested at scale before this code makes it to production, with a large number of reloads etc to ensure there are no issues since we won't be able to configure this for the time being so for this change, if you have 1000 tables and parallel reloads for all tables in 1 go on all servers, you end up 1000 x 1 x 50 (status object with only a Long object) ~ 30kB of memory (assuming every segment hosting every table!) This is because, we are storing 1 aggregate per job and no segment level items. Thus, I feel confident that this is not risky. However, we need to add configurability before we store segment level details in cache. ########## pinot-segment-local/src/main/java/org/apache/pinot/segment/local/utils/ServerReloadJobStatusCache.java: ########## @@ -0,0 +1,99 @@ +/** + * 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.pinot.segment.local.utils; + +import com.google.common.cache.Cache; +import com.google.common.cache.CacheBuilder; +import java.util.concurrent.TimeUnit; +import javax.annotation.Nullable; +import javax.annotation.concurrent.ThreadSafe; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import static java.util.Objects.requireNonNull; + + +/** + * In-memory cache for tracking reload job status on server side. + * Phase 1: Only tracks failure count per job. + * + * <p>Thread-safe for concurrent access. Uses Guava Cache with LRU eviction + * and time-based expiration. + */ +@ThreadSafe +public class ServerReloadJobStatusCache { + private static final Logger LOG = LoggerFactory.getLogger(ServerReloadJobStatusCache.class); + + private final Cache<String, ReloadJobStatus> _cache; + + /** + * Creates a cache with the given configuration. + * + */ + public ServerReloadJobStatusCache() { + final ServerReloadJobStatusCacheConfig config = new ServerReloadJobStatusCacheConfig(); Review Comment: yes. I think having some knobs like size of the reload cache in terms of # of jobs, or history length in days etc are helpful to keep the cache in check. or may be simply turning this off! hardcoding these params didn't seem the right approach. so, adding cache configurability will be part of the next set of upcoming changes. > do ensure this is well tested at scale before this code makes it to production, with a large number of reloads etc to ensure there are no issues since we won't be able to configure this for the time being so for this change, if you have 1000 tables and parallel reloads for all tables in 1 go on all servers, you end up 1000 x 1 x 50 (status object with only a Long object) ~50kB of memory (assuming every segment hosting every table!) This is because, we are storing 1 aggregate per job and no segment level items. Thus, I feel confident that this is not risky. However, we need to add configurability before we store segment level details in cache. -- 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] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
