We welcome the attention of the public number: Patricia learn Java , Java push to focus on areas of high-quality dry article! !
Website: https://www.exception.site/essay/create-tar-gz-by-java
One, background
Recently, Patricia primarily responsible for development work in the log table, and so on, what is the log table?
I only know that the concept stage, this time really fire, but the log table is used to doing in ?
Here Patricia try to log in layman's responsibilities under the table, before the log table to say, let's pull something else?
I believe we centralized logging platform ELK know some production environments, slightly more complex architectures, services are generally clustered deployment, so the log will be dispersed on each server, if a problem occurs, it will log you want to view very tedious, you need to log in to each server logs to find, because you are not sure which request was hit on the node. In addition, let the developers log in to view the log server itself is unsafe, careless implementation of rm -rf *the country?
By ELK, we can easily log into a collection ( Elasticsearchcluster) to analyze multi-dimensional.
But the deployment of high-performance, highly available ELK is the threshold, the business group wants to have the ability to quickly centralized log analysis, often need to go through the pre-technology research, test, foot pit in order to build up the platform.
Log in station's mission is to make the business line has this ability to quickly, simply complete fool of operation to access the log platform .
Smelly Hey! Having said that, with the theme of this article you what's the relationship?
Amount, Patricia enter this topic.
既然想统一管理日志，总得将这些分散的日志采集起来吧，那么，就需要一个日志采集器，Logstash 和 Filebeat 都有采集日志的能力，但是 Filebeat 相较于 Logstash 的笨重, 它更轻量级，几乎零占用服务器系统资源，这里我们选型 Filebeat。
业务组在日志平台完成相关接入流程后，平台会提供一个采集器包。接入方需要做的就是，下载这个采集器包并扔到指定服务器上，解压运行，即可开始采集日志，然后，就可以在日志平台的管控页面分析&搜索这些被收集的日志了。
这个
Filebeat采集器包里面，包含了采集日志文件路径，输出到Kafka集群，以及一些个性化的采集规则等等。
怎么样？是不是感觉很棒呢？
二、如何通过 Java 打包文件？
2.1 添加 Maven 依赖
<dependency>
<groupId>org.apache.commons</groupId>
<artifactId>commons-compress</artifactId>
<version>1.12</version>
</dependency>2.2 打包核心代码
通过 Apache compress 工具打包思路大致如下：
①：创建一个
FileOutputStream到输出文件（.tar.gz）文件。②：创建一个
GZIPOutputStream，用来包装FileOutputStream对象。③：创建一个
TarArchiveOutputStream，用来包装GZIPOutputStream对象。④：接着，读取文件夹中的所有文件。
⑤：如果是目录，则将其添加到
TarArchiveEntry。⑥：如果是文件，依然将其添加到
TarArchiveEntry中，然后还需将文件内容写入TarArchiveOutputStream中。
接下来，直接上代码：
import org.apache.commons.compress.archivers.tar.TarArchiveEntry;
import org.apache.commons.compress.archivers.tar.TarArchiveOutputStream;
import org.apache.commons.io.IOUtils;
import java.io.*;
import java.util.zip.GZIPOutputStream;
/**
* @author 犬小哈 (公众号： 小哈学Java)
* @date 2019-07-15
* @time 16:15
* @discription
**/
public class TarUtils {
/**
* 压缩
* @param sourceFolder 指定打包的源目录
* @param tarGzPath 指定目标 tar 包的位置
* @return
* @throws IOException
*/
public static void compress(String sourceFolder, String tarGzPath) throws IOException {
createTarFile(sourceFolder, tarGzPath);
}
private static void createTarFile(String sourceFolder, String tarGzPath) {
TarArchiveOutputStream tarOs = null;
try {
// 创建一个 FileOutputStream 到输出文件（.tar.gz）
FileOutputStream fos = new FileOutputStream(tarGzPath);
// 创建一个 GZIPOutputStream，用来包装 FileOutputStream 对象
GZIPOutputStream gos = new GZIPOutputStream(new BufferedOutputStream(fos));
// 创建一个 TarArchiveOutputStream，用来包装 GZIPOutputStream 对象
tarOs = new TarArchiveOutputStream(gos);
// 若不设置此模式，当文件名超过 100 个字节时会抛出异常，异常大致如下：
// is too long ( > 100 bytes)
// 具体可参考官方文档： http://commons.apache.org/proper/commons-compress/tar.html#Long_File_Names
tarOs.setLongFileMode(TarArchiveOutputStream.LONGFILE_POSIX);
addFilesToTarGZ(sourceFolder, "", tarOs);
} catch (IOException e) {
e.printStackTrace();
}finally{
try {
tarOs.close();
} catch (IOException e) {
e.printStackTrace();
}
}
}
public static void addFilesToTarGZ(String filePath, String parent, TarArchiveOutputStream tarArchive) throws IOException {
File file = new File(filePath);
// Create entry name relative to parent file path
String entryName = parent + file.getName();
// 添加 tar ArchiveEntry
tarArchive.putArchiveEntry(new TarArchiveEntry(file, entryName));
if (file.isFile()) {
FileInputStream fis = new FileInputStream(file);
BufferedInputStream bis = new BufferedInputStream(fis);
// 写入文件
IOUtils.copy(bis, tarArchive);
tarArchive.closeArchiveEntry();
bis.close();
} else if (file.isDirectory()) {
// 因为是个文件夹，无需写入内容，关闭即可
tarArchive.closeArchiveEntry();
// 读取文件夹下所有文件
for (File f : file.listFiles()) {
// 递归
addFilesToTarGZ(f.getAbsolutePath(), entryName + File.separator, tarArchive);
}
}
}
public static void main(String[] args) throws IOException {
// 测试一波，将 filebeat-7.1.0-linux-x86_64 打包成名为 filebeat-7.1.0-linux-x86_64.tar.gz 的 tar 包
compress("/Users/a123123/Work/filebeat-7.1.0-linux-x86_64", "/Users/a123123/Work/tmp_files/filebeat-7.1.0-linux-x86_64.tar.gz");
}
}至于，代码每行的作用，小伙伴们可以看代码注释，说的已经比较清楚了。
接下来，执行 main 方法，测试一下效果，看看打包是否成功：
生成采集器 tar.gz 包成功后，业务组只需将 tar.gz 下载下来，并扔到指定服务器，解压运行即可完成采集任务啦~
三、结语
本文主要还是介绍如何通过 Java 来完成打包功能，关于 ELK 相关的知识，小哈会在后续的文章中分享给大家，本文只是提及一下，欢迎小伙伴们持续关注哟，下期见~
四、Ref
欢迎关注微信公众号: 小哈学Java
更多干货文章，请关注笔者公众号: 小哈学Java (ID: xiaoha_java), 专注于分享Java领域干货文章, 不限于 BAT 面试题分享，Spring Boot, Spring Cloud, 高并发，JVM, 数据库，Docker 容器， ELK 等相关知识，另外，关注回复「资源」，即可获取全文最热的 Java 面试&架构学习资源哟~