• Feeds

  • Archive for the ‘data’ Category


    Memcached数据被踢(evictions>0)现象分析

    很多同学可能熟知Memcached的LRU淘汰算法,它是在slab内部进行的,如果所有空间都被slabs分配,即使另外一个slab里面有空位,仍然存在踢数据可能。你可以把slab理解为教室,如果你的教室满了,即使别的教室有空位你的教室也只能踢人才能进人。

    mc

    本文介绍的却是另外一种现象。今天监控发现线上一memcached发生数据被踢现象,用stats命令看evictions>0,因为以前也出现过此问题,后来对这个参数增加了一个监控,所以这次主动就发现了。由于给memcached分配的内存远大于业务存储数据所需内存,因此初步判断是“灵异现象”。

    第一步,netstat查看所有连接,排除是否被一些未规划的client使用,经排查后断定无此可能。

    第二步,用tcpdump抽样检查set的指令,排除是否有忘记设cache过期时间的client,初步检查所有典型的业务都有expire time。

    第三步,Google,未果

    第四步,看源代码,了解evictions计数器增加时的具体细节,oh, no…

    in items.c, memcached-1.2.8,

    125         for (search = tails[id]; tries > 0 && search != NULL; tries--, search=search->prev) {
    126             if (search->refcount == 0) {
    127                 if (search->exptime == 0 || search->exptime > current_time) {
    128                     itemstats[id].evicted++;
    129                     itemstats[id].evicted_time = current_time - search->time;
    130                     STATS_LOCK();
    131                     stats.evictions++;
    132                     STATS_UNLOCK();
    133                 }
    134                 do_item_unlink(search);
    135                 break;
    136             }
    137         }

    从源代码发现踢数据只判断一个条件,if (search->refcount == 0),这个refcount是多线程版本计数用,在当前服务器未启用多线程情况下,refcount应该始终为0,因此初步判断memcached是从访问队列尾部直接踢数据。

    为了证实想法,设计以下场景:

    1. 部署一个memcached测试环境,分配比较小的内存,比如8M
    2. 设置1条永远不过期的数据到memcached中,然后再get一次,这条数据后续应该存在LRU队尾。
    3. 每隔1S向memcached set(并get一次) 1,000条数据,过期时间设为3秒。
    4. 一段时间后,stats命令显示evictions=1

    按我以前的理解,第2步的数据是永远不会被踢的,因为有足够过期的数据空间可以给新来的数据用,LRU淘汰算法应该跳过没过期的数据,但结果证实这种判断是错误的。以上业务的服务器发生被踢的现象是由于保存了大量存活期短的key/value,且key是不重复的。另外又有一业务保存了小量不过期的数据,因此导致不过期的数据惨遭被挤到队列踢出。

    本来这个问题就告一段落了,但在写完这篇文章后,顺便又看了新一代memcached 1.4.1的源代码,很惊喜发现以下代码被增加。

    items.c, memcached 1.4.1

    107     /* do a quick check if we have any expired items in the tail.. */
    108     int tries = 50;
    109     item *search;
    110
    111     for (search = tails[id];
    112          tries > 0 && search != NULL;
    113          tries--, search=search->prev) {
    114         if (search->refcount == 0 &&
    115             (search->exptime != 0 && search->exptime < current_time)) {
    116             it = search;
    117             /* I don't want to actually free the object, just steal
    118              * the item to avoid to grab the slab mutex twice ;-)
    119              */
    120             it->refcount = 1;
    121             do_item_unlink(it);
    122             /* Initialize the item block: */
    123             it->slabs_clsid = 0;
    124             it->refcount = 0;
    125             break;
    126         }
    127     }

    重复进行上述测试,未发生evictions。

    9/8 Update: 注意到L108的tries=50没有?试想把测试第2步设置51条不过期数据到cache中,情况会怎样?因此新版的Memcached也同样存在本文描述问题。

    几条总结:

    • 过期的数据如果没被显式调用get,则也要占用空间。
    • 过期的不要和不过期的数据存在一起,否则不过期的可能被踢。
    • 从节约内存的角度考虑,即使数据会过期,也不要轻易使用随机字符串作为key,尽量使用定值如uid,这样占用空间的大小相对固定。
    • 估算空间大小时候请用slab size计算,不要按value长度去计算。
    • 不要把cache当作更快的key value store来用, cache不是storage。

    MemcacheDB, Tokyo Tyrant, Redis performance test

    I had tested the following key-value store for set() and get()

    1. Test environment

    1.1 Hardware/OS

    2 Linux boxes in a LAN, 1 server and 1 test client
    Linux Centos 5.2 64bit
    Intel(R) Xeon(R) CPU E5410  @ 2.33GHz (L2 cache: 6M), Quad-Core * 2
    8G memory
    SCSI disk (standalone disk, no other access)

    1.2 Software version

    db-4.7.25.tar.gz
    libevent-1.4.11-stable.tar.gz
    memcached-1.2.8.tar.gz
    memcachedb-1.2.1-beta.tar.gz
    redis-0.900_2.tar.gz
    tokyocabinet-1.4.9.tar.gz
    tokyotyrant-1.1.9.tar.gz

    1.3 Configuration

    Memcachedb startup parameter
    Test 100 bytes
    ./memcachedb -H /data5/kvtest/bdb/data -d -p 11212 -m 2048 -N -L 8192
    (Update: As mentioned by Steve, the 100-byte-test missed the -N paramter, so I added it and updated the data)
    Test 20k bytes
    ./memcachedb -H /data5/kvtest/mcdb/data -d -p 11212 -b 21000 -N -m 2048

    Tokyo Tyrant (Tokyo Cabinet) configuration
    Use default Tokyo Tyrant sbin/ttservctl
    use .tch database, hashtable database

    ulimsiz=”256m”
    sid=1
    dbname=”$basedir/casket.tch#bnum=50000000″ # default 1M is not enough!
    maxcon=”65536″
    retval=0

    Redis configuration
    timeout 300
    save 900 1
    save 300 10
    save 60 10000
    # no maxmemory settings

    1.4 Test client

    Client in Java, JDK1.6.0, 16 threads
    Use Memcached client java_memcached-release_2.0.1.jar
    JRedis client for Redis test, another JDBC-Redis has poor performance.

    2. Small data size test result

    Test 1, 1-5,000,000 as key, 100 bytes string value, do set, then get test, all get test has result.
    Request per second(mean)key-value-performance-1(Update)

    Store Write Read
    Memcached 55,989 50,974
    Memcachedb 25,583 35,260
    Tokyo Tyrant 42,988 46,238
    Redis 85,765 71,708

    Server Load Average

    Store Write Read
    Memcached 1.80, 1.53, 0.87 1.17, 1.16, 0.83
    MemcacheDB 1.44, 0.93, 0.64 4.35, 1.94, 1.05
    Tokyo Tyrant 3.70, 1.71, 1.14 2.98, 1.81, 1.26
    Redis 1.06, 0.32, 0.18 1.56, 1.00, 0.54

    3. Larger data size test result

    Test 2, 1-500,000 as key, 20k bytes string value, do set, then get test, all get test has result.
    Request per second(mean)
    (Aug 13 Update: fixed a bug on get() that read non-exist key)
    key-value-performance-2(update)

    Store Write Read
    Memcachedb 357 327
    Tokyo Tyrant 3,501 257
    Redis 1,542 957

    4. Some notes about the test

    When test Redis server, the memory goes up steadily, consumed all 8G and then use swap(and write speed slow down), after all memory and swap space is used, the client will get exceptions. So use Redis in a productive environment should limit to a small data size. It is another cache solution rather than a persistent storage. So compare Redis together with MemcacheDB/TC may not fair because Redis actually does not save data to disk during the test.

    Tokyo cabinet and memcachedb are very stable during heavy load, use very little memory in set test and less than physical memory in get test.

    MemcacheDB peformance is poor for write large data size(20k).

    The call response time was not monitored in this test.

    First...345