数据库
首页 > 数据库> > 分析redis key大小的几种方法

分析redis key大小的几种方法

作者:互联网

当redis被用作缓存时,有时我们希望了解key的大小分布,或者想知道哪些key占的空间比较大。本文提供了几种方法。

一. bigKeys

这是redis-cli自带的一个命令。对整个redis进行扫描,寻找较大的key。例:

redis-cli -h b.redis -p 1959 --bigkeys

输出:

# Scanning the entire keyspace to find biggest keys as well as
# average sizes per key type.  You can use -i 0.1 to sleep 0.1 sec
# per 100 SCAN commands (not usually needed).

[00.00%] Biggest hash   found so far 's_9329222' with 3 fields
[00.00%] Biggest string found so far 'url_http://mini.eastday.com/mobile/170722090206890.html?qid=sgllq&ch=east_sogou_push&pushid=13' with 8 bytes
[00.00%] Biggest string found so far 'foo' with 40 bytes
[00.00%] Biggest hash   found so far 's_9329084' with 4 fields
[00.23%] Biggest zset   found so far 'region_hot_菏泽地' with 625 members
[00.23%] Biggest zset   found so far 'region_hot_葫芦岛' with 914 members
[00.47%] Biggest string found so far 'top_notice_list' with 135193 bytes
[00.73%] Biggest zset   found so far 'region_hot_自贡' with 2092 members
[01.90%] Biggest hash   found so far 'uno_facet_2018-12-20' with 59 fields
[11.87%] Biggest zset   found so far 'region_hot_上海' with 2233 members
[27.05%] Biggest set    found so far 'blacklist_set_key' with 31832 members
[73.87%] Biggest string found so far 'PUSH_NEWS' with 3104237 bytes
[86.18%] Biggest zset   found so far 'region_hot_北京' with 2688 members

-------- summary -------

Sampled 4263 keys in the keyspace!
Total key length in bytes is 174847 (avg len 41.02)

Biggest string found 'PUSH_NEWS' has 3104237 bytes
Biggest    set found 'blacklist_set_key' has 31832 members
Biggest   hash found 'uno_facet_2018-12-20' has 59 fields
Biggest   zset found 'region_hot_北京' has 2688 members

1616 strings with 3771161 bytes (37.91% of keys, avg size 2333.64)
0 lists with 0 items (00.00% of keys, avg size 0.00)
1 sets with 31832 members (00.02% of keys, avg size 31832.00)
2353 hashs with 7792 fields (55.20% of keys, avg size 3.31)
293 zsets with 333670 members (06.87% of keys, avg size 1138.81)

说明:

  1. 该命令使用scan方式对key进行统计,所以使用时无需担心对redis造成阻塞。
  2. 输出大概分为两部分,summary之上的部分,只是显示了扫描的过程。summary部分给出了每种数据结构中最大的Key。
  3. 统计出的最大key只有string类型是以字节长度为衡量标准的。list,set,zset等都是以元素个数作为衡量标准,不能说明其占的内存就一定多。所以,如果你的Key主要以string类型存在,这种方法就比较适合。

更多关于bigkeys的说明可以参考这里

二. debug object key

redis的命令,可以查看某个key序列化后的长度。
例:

连接上redis后执行如下命令

b.redis:1959> hmset myhash k1 v1 k2 v2 k3 v3
OK
b.redis:1959> debug object myhash
Value at:0x7f005c6920a0 refcount:1 encoding:ziplist serializedlength:36 lru:3341677 lru_seconds_idle:2

关于输出的项的说明:

几个需要注意的问题

b.redis:1959> set str1 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
OK
b.redis:1959> set str2 abcdefghijklmnopqrstuvwxyz1234
OK
b.redis:1959> debug object str1
Value at:0x7f007c035b80 refcount:1 encoding:embstr serializedlength:12 lru:3342615 lru_seconds_idle:13
b.redis:1959> debug object str2
Value at:0x7f00654df400 refcount:1 encoding:embstr serializedlength:31 lru:3342622 lru_seconds_idle:7

两个字串的实际长度都是30, 但str1的serializedlength为12, str2的为31。

三. redis rdb tools

这是一个redis rdb file的分析工具,可以根据rdb file生成内存报告。

3.1 安装

需要python2.4以上版本和pip。

pip install rdbtools

3.2 生成内存报告

首先我们需要有一份rdb文件,如果你在配置中开启了rdb,那么redis会自动生成rdb文件。如果没有,可以手动执行bgsave。如果是线上机器,执行时要考虑机器负载等问题。拿到rdb文件后,我们就可以生成内存报告了。命令如下:

rdb -c memory file 

例:

rdb -c memory /tmp/dump.rdb 
database,type,key,size_in_bytes,encoding,num_elements,len_largest_element,expiry
0,hash,data:index_flow_yingshi,10492,hashtable,1,8992,2019-01-14T08:20:10.236000
0,hash,data:index_movie,22068,hashtable,7,2896,2019-01-14T07:29:19.685000
0,string,block:index_module_novel,8296,string,7694,7694,2019-01-13T00:27:46.128000
0,string,block:index_bottom_baike_aikan,8296,string,7632,7632,2019-01-14T02:27:11.850000
0,string,block:index_bottom_tools,5224,string,4549,4549,2019-01-13T01:02:09.171000
0,string,block:index_module_travel,7272,string,6408,6408,2019-01-13T00:43:39.478000
...

输出了db,数据类型,key, 大小, 编码等多列信息。至于分析数据,你可以用shell,也可以保存成csv用excel排序,或者干脆存到db里,想怎么排怎么排。

如果只要知道最大的N个key, 可以使用-l选项。例:

[@sjs_73_171 ~]$ rdb -c memory -l 3 /tmp/dump.rdb  
database,type,key,size_in_bytes,encoding,num_elements,len_largest_element,expiry
0,hash,city_tong,724236,hashtable,3113,216,2019-01-14T01:10:59.407000
0,hash,iplocsearch,406292,hashtable,383,180190,2019-01-30T05:37:56.082000
0,hash,weather_tong3,583844,hashtable,319,1658,2019-01-07T10:22:33.742000

3.3 查看单个key

如果我们只需要查询单个key所使用的内存可以不必依赖rdb file, 使用redis-memory-for-key命令即可。
例:

[@sjs_73_171 WEB-INF]$ redis-memory-for-key -s b.redis -p 1959 myhash
Key                             myhash
Bytes                           83
Type                            hash
Encoding                        ziplist
Number of Elements              3
Length of Largest Element       2

[@sjs_73_171 WEB-INF]$ redis-memory-for-key -s b.redis -p 1959 str1
Key                             str1
Bytes                           80
Type                            string

[@sjs_73_171 WEB-INF]$ redis-memory-for-key -s b.redis -p 1959 str2
Key                             str2
Bytes                           80
Type                            string

3.4 更多

rdb --help

也可以查看git上的帮助文档

四. 总结

标签:string,redis,几种,key,rdb,Biggest,found
来源: https://www.cnblogs.com/duanweishi/p/14148943.html