在提交作业时，避免每次上传外部JAR的Flink配置，可以使用Flink的集群模式来实现。下面是一个示例代码：

import org.apache.flink.api.java.ExecutionEnvironment;
import org.apache.flink.api.java.utils.ParameterTool;
import org.apache.flink.client.program.ClusterClient;
import org.apache.flink.client.program.PackagedProgram;
import org.apache.flink.client.program.ProgramInvocationException;
import org.apache.flink.client.program.ProgramMissingJobException;
import org.apache.flink.configuration.Configuration;
import org.apache.flink.runtime.client.JobExecutionException;
import org.apache.flink.runtime.jobgraph.JobGraph;
import org.apache.flink.yarn.client.YarnClusterClientFactory;

import java.io.File;
import java.net.URL;
import java.util.Arrays;
import java.util.List;

public class FlinkJobSubmitter {

    public static void main(String[] args) throws Exception {
        // 设置Flink配置
        Configuration flinkConfig = new Configuration();
        flinkConfig.setInteger("parallelism", 4);

        // 创建一个Flink集群客户端
        YarnClusterClientFactory clusterClientFactory = new YarnClusterClientFactory();
        ClusterClient clusterClient = clusterClientFactory.createClusterClient(flinkConfig);

        // 创建Flink执行环境
        ExecutionEnvironment env = ExecutionEnvironment.createRemoteEnvironment(clusterClient);

        // 添加外部JAR依赖
        List jarFiles = Arrays.asList(
                new File("/path/to/external.jar").toURI().toURL()
        );
        env.registerExternalJars(jarFiles.toArray(new URL[0]));

        // 使用ParameterTool解析命令行参数
        ParameterTool params = ParameterTool.fromArgs(args);

        // 设置作业参数
        env.getConfig().setGlobalJobParameters(params);

        // 提交作业
        try {
            PackagedProgram program = new PackagedProgram(new File("/path/to/job.jar"));
            JobGraph jobGraph = program.getJobGraph();
            clusterClient.runDetached(jobGraph);
        } catch (ProgramMissingJobException e) {
            System.err.println("No Flink job found in JAR.");
        } catch (ProgramInvocationException e) {
            System.err.println("Failed to invoke Flink job.");
            e.printStackTrace();
        } catch (JobExecutionException e) {
            System.err.println("Failed to execute Flink job.");
            e.printStackTrace();
        }

        // 关闭集群客户端
        clusterClient.shutdown();
    }
}

上述代码使用了YarnClusterClient来创建一个Flink集群客户端，并通过ExecutionEnvironment创建了一个远程执行环境。在提交作业之前，通过env.registerExternalJars()方法添加了外部JAR依赖。然后，使用ParameterTool解析命令行参数，并使用PackagedProgram和JobGraph来构建和提交作业。最后，关闭集群客户端。

在实际使用中，你需要将代码中的/path/to/external.jar和/path/to/job.jar替换为你实际的外部JAR文件路径。