May I ask why a timeout of 6 seconds has been configured, why the job has not given up resources and is still blocking, and other requests cannot come in?

jonathan.lukas · June 30, 2023, 6:46am

this is already helpful. But I am still convinced that a view on the BPMN diagrams will allow for more insights regarding your issue.

Jonathan

bugbugmaker · July 5, 2023, 7:32am

Will the Zeebe client only actively attempt to pull tasks after one of the tasks is completed or fails? If there is only one task in a certain pull, the jobhandler executing this task happens to be blocked or time-consuming. Before this task fails or is completed, the zeebe client will not actively pull new tasks. Doesn’t this lead to thread redundancy? Can other newly created tasks not be processed in a timely manner?

jonathan.lukas · July 7, 2023, 4:45am

Hello @bugbugmaker ,

here you can see that the tasks are polled from the engine in a non-blocking manner:

github.com

camunda/zeebe/blob/main/clients/java/src/main/java/io/camunda/zeebe/client/impl/worker/JobWorkerImpl.java

/*
 * Copyright © 2017 camunda services GmbH (info@camunda.com)
 *
 * Licensed 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 io.camunda.zeebe.client.impl.worker;

import io.camunda.zeebe.client.api.response.ActivatedJob;
import io.camunda.zeebe.client.api.worker.BackoffSupplier;
import io.camunda.zeebe.client.api.worker.JobWorker;

This file has been truncated. show original

The client has a defined max amount of jobs it polls for. The poll will continue until this amount of active jobs is reached. As long as the amount of remaining jobs is higher than 30% of the max active jobs, the worker will not poll for new jobs.

I hope this helps

Jonathan

bugbugmaker · July 11, 2023, 8:14am

What I want to ask is why the timeout configuration did not take effect, or what is the actual meaning, and how to use timeout.
The result of my testing is that the configured 6 seconds do not take effect, as it seems to be executed by other workers every 30 seconds

jonathan.lukas · July 19, 2023, 2:11pm

Hello @bugbugmaker ,

this is an interesting finding. For me, the timeout always works as expected. Could you increase the maxJobsActive parameter to check whether the execution is deferred by the client as long as there is a job execution still active of the type?

Jonathan