serverWatches.forEach( sw -> { String serverId = sw.getId(); String ip = sw.getIp(); try { saveSystemMonitor(serverId, ip, time, plmxs_url); } catch (Exception e) { log.error("保存cpu使用率|内存使用率失败:{}", e); } try { saveDiskIoDetail(serverId, ip, time, plmxs_url); } catch (Exception e) { log.error("保存磁盘IO失败:{}", e); } try { saveNetIoReceive(serverId, ip, time, plmxs_url); } catch (Exception e) { log.error("保存网络IO-receive失败:{}", e); } try { saveNetIoTransmit(serverId, ip, time, plmxs_url); } catch (Exception e) { log.error("保存网络IO-transmit失败:{}", e); } } );

可以将每个保存操作都放入一个线程中进行执行，这样可以提高并发处理能力，代码如下所示：

serverWatches.parallelStream().forEach(sw -> {
    String serverId = sw.getId();
    String ip = sw.getIp();
    ExecutorService executor = Executors.newFixedThreadPool(4);
    executor.execute(() -> {
        try {
            saveSystemMonitor(serverId, ip, time, plmxs_url);
        } catch (Exception e) {
            log.error("保存cpu使用率|内存使用率失败:{}", e);
        }
    });
    executor.execute(() -> {
        try {
            saveDiskIoDetail(serverId, ip, time, plmxs_url);
        } catch (Exception e) {
            log.error("保存磁盘IO失败:{}", e);
        }
    });
    executor.execute(() -> {
        try {
            saveNetIoReceive(serverId, ip, time, plmxs_url);
        } catch (Exception e) {
            log.error("保存网络IO-receive失败:{}", e);
        }
    });
    executor.execute(() -> {
        try {
            saveNetIoTransmit(serverId, ip, time, plmxs_url);
        } catch (Exception e) {
            log.error("保存网络IO-transmit失败:{}", e);
        }
    });
    executor.shutdown();
});

在上面的代码中，我们使用了 Java 自带的线程池进行多线程处理，其中 `newFixedThreadPool(4)` 表示创建一个固定大小为 4 的线程池，每个保存操作都会被分配到其中一个线程中进行执行。最后，我们需要在代码结束前调用 `executor.shutdown()` 来关闭线程池。