rhauch commented on a change in pull request #11797: URL: https://github.com/apache/kafka/pull/11797#discussion_r819851789
########## File path: connect/runtime/src/main/java/org/apache/kafka/connect/util/RetryUtil.java ########## @@ -0,0 +1,107 @@ +/* + * 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.kafka.connect.util; + +import org.apache.kafka.common.errors.RetriableException; +import org.apache.kafka.common.errors.WakeupException; +import org.apache.kafka.common.utils.Utils; +import org.apache.kafka.connect.errors.ConnectException; +import org.slf4j.Logger; +import org.slf4j.LoggerFactory; + +import java.time.Duration; +import java.util.Optional; +import java.util.concurrent.Callable; +import java.util.function.Supplier; + +public class RetryUtil { + private static final Logger log = LoggerFactory.getLogger(RetryUtil.class); + + /** + * The method executes the callable at least once, optionally retrying the callable if + * {@link org.apache.kafka.connect.errors.RetriableException} is being thrown. If timeout is exhausted, + * then the last exception is wrapped with a {@link org.apache.kafka.connect.errors.ConnectException} and thrown. + * + * <p>{@code description} supplies the message that indicates the purpose of the callable since the message will + * be used for logging. For example, "list offsets". If the supplier is null or the supplied string is + * null, {@code callable} will be used as the default string. + * + * <p>The task will be executed at least once. No retries will be performed + * if {@code timeoutDuration} is 0 or negative, or if {@code timeoutDuration} is less than {@code retryBackoffMs}. + * + * <p>A {@code retryBackoffMs} that is negative or zero will result in no delays between retries. + * + * @param callable the function to execute + * @param description supplier that provides custom message for logging purpose + * @param timeoutDuration timeout duration; must not be null + * @param retryBackoffMs the number of milliseconds to delay upon receiving a + * {@link org.apache.kafka.connect.errors.RetriableException} before retrying again + * @throws ConnectException If the task exhausted all the retries + */ + public static <T> T retryUntilTimeout(Callable<T> callable, Supplier<String> description, Duration timeoutDuration, long retryBackoffMs) throws Exception { + // if null supplier or string is provided, the message will be default to "callabe" + String descriptionStr = Optional.ofNullable(description) + .map(Supplier::get) + .orElse("callable"); + + // handling null duration + long timeoutMs = Optional.ofNullable(timeoutDuration) + .map(Duration::toMillis) + .orElse(0L); + + if (retryBackoffMs >= timeoutMs) { + log.debug("Executing {} only once, since retryBackoffMs={} is larger than total timeoutMs={}", + descriptionStr, retryBackoffMs, timeoutMs); + } + + if (retryBackoffMs < 0) { + log.debug("Invalid retryBackoffMs, must be non-negative but got {}. 0 will be used instead", + retryBackoffMs); + retryBackoffMs = 0; + } + + if (timeoutMs <= 0 || retryBackoffMs >= timeoutMs) { + // no retry + return callable.call(); + } + + long end = System.currentTimeMillis() + timeoutMs; + int attempt = 0; + Throwable lastError = null; + while (System.currentTimeMillis() <= end) { Review comment: I think there's an edge case where `timeoutMs` is positive but small enough that the condition on line 77 is not met but the while loop on line 85 is not satisfied because the end time has already passed. In this edge case, we might not call the callable function (even once). One option is to change the while loop to be a do-while loop so that we always go through one loop. Another option is to compute the remaining time before line 77 and not update it before the while loop. Either would work, but one of the options may require fewer duplicated lines. -- 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: jira-unsubscr...@kafka.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org
