Elasticsearch数据备份与恢复(基于HDFS)

1.(所有机子上)安装es hdfs仓库插件repository-hdfs

# repository-hdfs一定要和es版本匹配
# 在线
bin/elasticsearch-plugin install repository-hdfs

# 离线
bin/elasticsearch-plugin install file:///xxx/repository-hdfs-x.x.x.zip.zip

2. (所有机子上)设置ES-HDFS仓库安全策略

plugins/repository-hdfs/plugin-security.policy追加

permission java.lang.RuntimePermission "accessDeclaredMembers";
permission java.lang.RuntimePermission "getClassLoader";
permission java.lang.RuntimePermission "shutdownHooks";
permission java.lang.reflect.ReflectPermission "suppressAccessChecks";
permission javax.security.auth.AuthPermission "doAs";
permission javax.security.auth.AuthPermission "getSubject";
permission javax.security.auth.AuthPermission "modifyPrivateCredentials";
permission java.security.AllPermission;
permission java.util.PropertyPermission "*", "read,write";
permission javax.security.auth.PrivateCredentialPermission "org.apache.hadoop.security.Credentials * \"*\"", "read";

config/jvm.options追加

Djava.security.policy=file:////xxx/plugins/repository-hdfs/plugin-security.policy

3.(所有机子上)设置JAVA安全管理器

$JAVA_HOME/jre/lib/security/java.policy追加

permission java.security.AllPermission;

重启Es集群

4. (一台机子上)创建HDFS仓库

# 创建远程仓库,目录为/user/xxx/es/repository/es_hdfs_repository
# 在HDFS 50070查看
# uri需要是active的namenode节点
curl -XPUT 'http://host:esPort/_snapshot/es_hdfs_repository' -d '
{
    "type":"hdfs",
    "settings":{
        "uri":"hdfs://host:hdfsPort/",
        "path":"es/repository/es_hdfs_repository"
    }
}'

5. (一台机子上)备份

# 对所有open的index进行备份,创建一个snapshot_1的备份
# wait_for_completion会堵塞到备份完成
curl -XPUT 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1?wait_for_completion=true'

# 指定索引备份
curl -XPUT 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1?wait_for_completion=true' -d '
{
    "indices":["index1","index2"],
    "ignore_unavailable":true,
    "include_global_state":false,
    "partial":true
}'

# 查看备份进度和结果
curl -XGET 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1?pretty'

6.(一台机子上)恢复

# 恢复指定索引
#include_global_state=true 集群状态也备份
#ignore_unavailable 部分索引不存在时也通过,否则认为不成功
#partial=true 部分失效也可以通过,否则认为不成功
curl -XPOST 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1/_restore?wait_for_completion=true' -d '
{
    "indices":["redisindex"],
    "ignore_unavailable":true,
    "include_global_state":false,
    "partial":true
}'

# 恢复snapshot_1全部索引
curl -XPOST 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1/_restore?wait_for_completion=true'

#恢复进度和结果
curl -XGET 'http://host:esPort/index1/_recovery?pretty'

Tips:

恢复之前最好先把所有close,防止恢复期间还有新数据产生

如果是不存在的索引,不在考虑范围

7.其他

# 关闭索引
curl -XPOST 'hhtp://host:esPort/eventindex/_close'

# 打开索引
curl -XPOST 'http://host:esPort/eventindex/_open'

# 删除备份
# 备份是增量备份的,所以不要轻易手动删,否则会出现索引损坏而无法恢复
# 正确的姿势
curl -XDELETE 'http://host:esPort/_snapshot/es_hdfs_repository/snapshot_1'