在java里向spark请求处理

打印spark处理失败的日志

SparkLauncher launcher = sparkJobUtil.buildSparkLauncher(feedConfig, appName, params);
SparkAppHandle handler = launcher.startApplication();
int exitCode = -1;
while (handler.getState() == null || !handler.getState().isFinal()) {
    if (handler.getState() != null) {
        if (handler.getState().equals(SparkAppHandle.State.FINISHED)) {
            log.info("instance id {} App id: {} is finished ", omFeed.getInstanceId(), handler.getAppId());
            exitCode = 0;
        }
    }
}
if (handler.getError().isPresent()) {
    log.error("instance id  {} app id  fail fail ", omFeed.getInstanceId(), handler.getAppId());
    handler.getError().get().printStackTrace();
}else{
    log.error("instance id  {} app id  run success ", omFeed.getInstanceId(), handler.getAppId());
    exitCode = 0 ;
}

标签：omFeed,getState,java,请求,getInstanceId,handler,spark,id,log
来源： https://www.cnblogs.com/ivyJ/p/16375855.html