公司服务器上安装了contly，是一个开源的node.js项目，用于统计手机app使用情况，后端数据储存使用的mongodb，使用的时候经常发现mongodb占用cpu非常高，打到了210%的爆表值
top - 13:42:39 up 308 days, 23:01, 2 users, load average: 2.84, 2.96, 2.93 Tasks: 209 total, 1 running, 208 sleeping, 0 stopped, 0 zombie %Cpu(s): 59.4 us, 2.7 sy, 0.0 ni, 36.9 id, 0.2 wa, 0.7 hi, 0.0 si, 0.0 st KiB Mem: 8173524 total, 6095460 used, 2078064 free, 133752 buffers KiB Swap: 1048572 total, 750536 used, 298036 free. 3480688 cached Mem PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 16870 root 20 0 48.452g 2.643g 2.490g S 210.5 33.9 38607:41 mongod 28508 root 20 0 891316 110272 5976 S 5.0 1.3 10:07.08 nodejs 28589 root 20 0 893736 114044 5972 S 5.0 1.4 13:19.13 nodejs 28566 root 20 0 959004 113128 5976 S 4.0 1.4 11:16.49 nodejs 28670 root 20 0 1189932 383820 5980 S 4.0 4.7 26:00.97 nodejs 16367 www-data 20 0 86960 2952 996 S 1.0 0.0 237:47.74 nginx 16368 www-data 20 0 87028 3212 996 S 1.0 0.0 235:54.43 nginx 16370 www-data 20 0 86964 2952 996 S 1.0 0.0 245:46.74 nginx 16371 www-data 20 0 86960 2948 996 S 1.0 0.0 246:40.89 nginx 1154 root 20 0 23792 1764 1144 R 0.7 0.0 0:00.23 top 7 root 20 0 0 0 0 S 0.3 0.0 1053:48 rcu_sched 8 root 20 0 0 0 0 S 0.3 0.0 229:02.41 rcuos/0 6559 root 20 0 3957096 217820 5668 S 0.3 2.7 230:37.95 jsvc 1 root 20 0 37348 3704 976 S 0.0 0.0 0:19.37 init 2 root 20 0 0 0 0 S 0.0 0.0 0:02.51 kthreadd 3 root 20 0 0 0 0 S 0.0 0.0 68:42.78 ksoftirqd/0 5 root 0 -20 0 0 0 S 0.0 0.0 0:00.00 kworker/0:0H 9 root 20 0 0 0 0 S 0.0 0.0 225:56.62 rcuos/1 10 root 20 0 0 0 0 S 0.0 0.0 217:26.86 rcuos/2 11 root 20 0 0 0 0 S 0.0 0.0 228:57.90 rcuos/3 12 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/4 13 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/5 14 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/6 15 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/7 16 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/8 17 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/9 18 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/10 19 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/11 20 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/12 21 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/13 22 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/14 23 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/15 24 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/16 25 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/17 26 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/18 27 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/19 28 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/20 29 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/21 30 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/22 31 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/23 32 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/24 33 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/25 34 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/26 35 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/27 36 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/28 37 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/29 38 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/30 39 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuos/31 40 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcu_bh 41 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/0 42 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/1 43 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/2 44 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/3 45 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/4 46 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/5 47 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/6 48 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/7 49 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/8 50 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/9 51 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/10 52 root 20 0 0 0 0 S 0.0 0.0 0:00.00 rcuob/11
仔细查看是中断很多，命令是pid -w 1
root@localhost:~# pidstat -w 1 Linux 3.13.0-24-generic (localhost) 07/14/2017 _x86_64_ (4 CPU) 01:44:41 PM UID PID cswch/s nvcswch/s Command 01:44:42 PM 0 7 260.40 0.00 rcu_sched 01:44:42 PM 0 8 78.22 0.00 rcuos/0 01:44:42 PM 0 9 49.50 0.00 rcuos/1 01:44:42 PM 0 10 88.12 0.00 rcuos/2 01:44:42 PM 0 11 96.04 0.00 rcuos/3 01:44:42 PM 0 74 0.99 0.00 watchdog/0 01:44:42 PM 0 75 0.99 0.00 watchdog/1 01:44:42 PM 0 77 7.92 0.00 ksoftirqd/1 01:44:42 PM 0 80 0.99 0.00 watchdog/2 01:44:42 PM 0 82 39.60 0.00 ksoftirqd/2 01:44:42 PM 0 85 0.99 0.00 watchdog/3 01:44:42 PM 0 87 9.90 0.00 ksoftirqd/3 01:44:42 PM 0 1103 7.92 0.00 kworker/u64:2 01:44:42 PM 0 1213 0.99 4.95 pidstat 01:44:42 PM 0 1265 0.99 0.00 supervisord 01:44:42 PM 0 1279 9.90 0.00 vmtoolsd 01:44:42 PM 0 1373 0.99 0.00 fail2ban-server 01:44:42 PM 0 1954 18.81 0.00 xfsaild/sdb1 01:44:42 PM 0 2132 9.90 0.00 kworker/1:1H 01:44:42 PM 0 6800 0.99 0.00 kworker/0:1 01:44:42 PM 0 6807 0.99 0.00 kworker/2:2 01:44:42 PM 1000 14269 0.99 0.00 zabbix_agentd 01:44:42 PM 1000 14273 0.99 0.00 zabbix_agentd 01:44:42 PM 33 16367 154.46 0.00 nginx 01:44:42 PM 33 16368 178.22 0.00 nginx 01:44:42 PM 33 16370 166.34 0.00 nginx 01:44:42 PM 33 16371 168.32 0.00 nginx 01:44:42 PM 0 16870 100.00 0.00 mongod 01:44:42 PM 0 28498 0.99 0.00 supervisord 01:44:42 PM 0 28508 126.73 39.60 nodejs 01:44:42 PM 0 28566 141.58 46.53 nodejs 01:44:42 PM 0 28589 137.62 110.89 nodejs 01:44:42 PM 0 28670 123.76 96.04 nodejs 01:44:42 PM 0 28716 6.93 0.00 kworker/3:1 01:44:42 PM 0 28732 0.99 0.00 kworker/1:0
vmstat
root@localhost:~# vmstat 1 procs -----------memory---------- ---swap-- -----io---- -system-- ------cpu----- r b swpd free buff cache si so bi bo in cs us sy id wa st 3 1 750536 2089980 133864 3482948 0 0 8 738 0 0 14 4 81 1 0 3 0 750536 2089932 133864 3482948 0 0 0 2062 6793 9336 67 5 28 0 0 3 0 750536 2086820 133864 3482952 0 0 0 1767 6125 8760 72 5 23 0 0 4 0 750536 2084232 133864 3482952 0 0 0 1611 6227 8686 60 4 37 0 0 3 0 750536 2088264 133864 3482952 0 0 0 140 5974 8556 65 4 31 0 0 4 0 750536 2083864 133864 3482956 0 0 0 2635 9100 10589 78 5 17 0 0 2 0 750536 2085624 133864 3482916 0 0 0 168 8152 11602 63 5 33 0 0
中断和上下文切换多，是因为nodejs一直在被从sleep中唤醒，实际上是countly要接受手机端返回来的数据，不断被IO唤醒，每次唤醒都要上下文切换，所以表象是上下文频繁切换，实质上是counly负担比较重。
不过这个貌似和mongodb没有关系，只能说明countly的负担比较重，一直在写入mongodb，所以现在mongodb的负担也比较大。其实io算不上高，cpu的iowait也不多，硬盘写入并不是瓶颈，可能是mongodb本身的问题
查看mongostat
root@localhost:~# mongostat connected to: 127.0.0.1 insert query update delete getmore command flushes mapped vsize res faults locked db idx miss % qr|qw ar|aw netIn netOut conn time 13 171 424 2 0 15|0 0 24g 48.5g 2.64g 0 countly:11.9% 0 0|0 3|0 102k 67k 12 13:51:04 9 225 1358 105 0 13|0 0 24g 48.5g 2.65g 0 countly:6.5% 0 0|0 0|1 255k 177k 12 13:51:05 218 967 3056 11 0 82|0 0 24g 48.5g 2.63g 0 countly:20.7% 0 0|0 2|0 807k 965k 12 13:51:06 19 158 315 2 0 12|0 0 24g 48.5g 2.65g 0 countly:1.7% 0 0|0 1|0 88k 92k 12 13:51:07 17 132 270 2 0 8|0 0 24g 48.5g 2.58g 0 countly:2.1% 0 0|0 2|0 77k 115k 12 13:51:08 14 151 433 2 0 11|0 0 24g 48.5g 2.64g 0 countly:2.2% 0 0|0 3|0 106k 89k 12 13:51:09 22 181 335 3 0 17|0 0 24g 48.5g 2.66g 0 countly:7.9% 0 0|0 3|0 96k 105k 12 13:51:10 36 230 765 3 0 8|0 0 24g 48.5g 2.66g 0 countly:3.6% 0 1|0 0|1 184k 224k 12 13:51:11 122 567 1580 1 0 56|0 0 24g 48.5g 2.67g 0 countly:16.4% 0 0|0 2|0 440k 559k 12 13:51:12 8 124 286 4 0 8|0 0 24g 48.5g 2.65g 0 countly:2.0% 0 0|0 2|0 67k 61k 12 13:51:13 insert query update delete getmore command flushes mapped vsize res faults locked db idx miss % qr|qw ar|aw netIn netOut conn time 21 193 450 1 0 18|0 0 24g 48.5g 2.66g 0 countly:5.0% 0 0|0 3|0 119k 126k 12 13:51:14 156 774 950 1 0 67|0 0 24g 48.5g 2.65g 0 countly:18.5% 0 0|0 3|0 414k 623k 12 13:51:15 7 69 194 2 0 12|0 0 24g 48.5g 2.65g 0 countly:1.4% 0 0|0 3|0 49k 40k 12 13:51:16 33 280 525 2 0 12|0 0 24g 48.5g 2.64g 0 countly:4.8% 0 0|0 3|0 145k 187k 12 13:51:17 18 201 563 5 0 17|0 0 24g 48.5g 2.66g 0 countly:5.5% 0 0|0 3|0 141k 126k 12 13:51:18 10 144 334 2 0 15|0 0 24g 48.5g 2.65g 0 countly:4.4% 0 0|0 4|0 84k 103k 12 13:51:19 14 210 525 4 0 19|0 0 24g 48.5g 2.63g 0 countly:5.8% 0 0|0 1|0 125k 96k 12 13:51:20 563 1959 2259 2 0 222|0 0 24g 48.5g 2.61g 0 countly:49.2% 0 0|0 1|0 1m 2m 12 13:51:21 227 692 1326 2 0 60|0 0 24g 48.5g 2.68g 3 countly:14.3% 0 0|0 2|0 647k 652k 12 13:51:22 17 146 315 1 0 8|0 0 24g 48.5g 2.65g 0 countly:3.5% 0 2|0 0|1 87k 95k 12 13:51:23 insert query update delete getmore command flushes mapped vsize res faults locked db idx miss % qr|qw ar|aw netIn netOut conn time 23 166 376 1 0 10|0 0 24g 48.5g 2.65g 0 countly:7.4% 0 0|0 2|0 104k 104k 12 13:51:24 18 113 212 *0 0 7|0 0 24g 48.5g 2.64g 0 countly:12.6% 0 0|0 2|0 65k 75k 12 13:51:25 13 177 375 *0 0 19|0 0 24g 48.5g 2.64g 0 countly:2.4% 0 3|0 0|1 93k 110k 12 13:51:26 20 147 299 5 0 5|0 0 24g 48.5g 2.63g 0 countly:1.9% 0 0|0 2|0 85k 103k 12 13:51:27 40 280 476 2 0 15|0 1 24g 48.5g 2.63g 0 countly:2.9% 0 0|0 1|0 157k 247k 12 13:51:28 7 108 263 2 0 5|0 0 24g 48.5g 2.64g 0 countly:1.6% 0 0|0 2|0 62k 78k 12 13:51:29 17 181 394 *0 0 8|0 0 24g 48.5g 2.59g 0 countly:5.4% 0 0|0 2|0 101k 147k 12 13:51:30 17 152 350 1 0 12|0 0 24g 48.5g 2.63g 0 countly:3.7% 0 0|0 1|0 90k 80k 12 13:51:31 14 123 261 4 0 11|0 0 24g 48.5g 2.58g 0 countly:4.0% 0 2|0 0|1 72k 108k 12 13:51:32 184 741 1220 21 0 84|0 0 24g 48.5g 2.67g 0 countly:25.3% 0 0|0 4|0 492k 654k 12 13:51:33
mongodb的锁很多，怀疑是因为写入慢，一直在等待，浪费了大量的cpu时间
进入到mongodb中查看
root@localhost:~# mongo MongoDB shell version: 2.4.14 connecting to: test > db.serverStatus() { "host" : "localhost", "version" : "2.4.14", "process" : "mongod", "pid" : 16870, "uptime" : 1551490, "uptimeMillis" : NumberLong(1551489627), "uptimeEstimate" : 1537152, "localTime" : ISODate("2017-07-14T06:12:34.348Z"), "asserts" : { "regular" : 0, "warning" : 0, "msg" : 0, "user" : 1, "rollovers" : 0 }, "backgroundFlushing" : { "flushes" : 25858, "total_ms" : 6413133, "average_ms" : 248.01349679016164, "last_ms" : 111, "last_finished" : ISODate("2017-07-14T06:12:27.580Z") }, "connections" : { "current" : 12, "available" : 19988, "totalCreated" : NumberLong(660) }, "cursors" : { "totalOpen" : 2, "clientCursors_size" : 2, "timedOut" : 125, "totalNoTimeout" : 2 }, "dur" : { "commits" : 29, "journaledMB" : 0.28672, "writeToDataFilesMB" : 0.272729, "compression" : 0.9344831856907262, "commitsInWriteLock" : 0, "earlyCommits" : 0, "timeMs" : { "dt" : 3071, "prepLogBuffer" : 0, "writeToJournal" : 49, "writeToDataFiles" : 4, "remapPrivateView" : 5 } }, "extra_info" : { "note" : "fields vary by platform", "heap_usage_bytes" : 174045352, "page_faults" : 37855 }, "globalLock" : { "totalTime" : NumberLong("1551489627000"), "lockTime" : NumberLong("28450005021"), "currentQueue" : { "total" : 0, "readers" : 0, "writers" : 0 }, "activeClients" : { "total" : 2, "readers" : 2, "writers" : 0 } }, "indexCounters" : { "accesses" : 2559023213, "hits" : 2559023356, "misses" : 0, "resets" : 0, "missRatio" : 0 }, "locks" : { "." : { "timeLockedMicros" : { "R" : NumberLong(1631854144), "W" : NumberLong("28450005021") }, "timeAcquiringMicros" : { "R" : NumberLong(1839692044), "W" : NumberLong(269051643) } }, "admin" : { "timeLockedMicros" : { }, "timeAcquiringMicros" : { } }, "local" : { "timeLockedMicros" : { "r" : NumberLong(2235967), "w" : NumberLong(0) }, "timeAcquiringMicros" : { "r" : NumberLong(856059), "w" : NumberLong(0) } }, "countly" : { "timeLockedMicros" : { "r" : NumberLong("4052844846798"), "w" : NumberLong("77669514748") }, "timeAcquiringMicros" : { "r" : NumberLong("2069545770119"), "w" : NumberLong("82155316804") } }, "countly_drill" : { "timeLockedMicros" : { "r" : NumberLong(874145401), "w" : NumberLong("19474885608") }, "timeAcquiringMicros" : { "r" : NumberLong(271347249), "w" : NumberLong(1475646788) } }, "test" : { "timeLockedMicros" : { "r" : NumberLong(68640), "w" : NumberLong(79192) }, "timeAcquiringMicros" : { "r" : NumberLong(146), "w" : NumberLong(26) } } }, "network" : { "bytesIn" : 233060269131, "bytesOut" : 258427384819, "numRequests" : 1384182586 }, "opcounters" : { "insert" : 43661297, "query" : 434433635, "update" : 877502004, "delete" : 4880964, "getmore" : 1023, "command" : 26378757 }, "opcountersRepl" : { "insert" : 0, "query" : 0, "update" : 0, "delete" : 0, "getmore" : 0, "command" : 0 }, "recordStats" : { "accessesNotInMemory" : 8466, "pageFaultExceptionsThrown" : 3696, "countly" : { "accessesNotInMemory" : 1231, "pageFaultExceptionsThrown" : 221 }, "countly_drill" : { "accessesNotInMemory" : 7235, "pageFaultExceptionsThrown" : 3475 }, "local" : { "accessesNotInMemory" : 0, "pageFaultExceptionsThrown" : 0 }, "test" : { "accessesNotInMemory" : 0, "pageFaultExceptionsThrown" : 0 } }, "writeBacksQueued" : false, "mem" : { "bits" : 64, "resident" : 2692, "virtual" : 49774, "supported" : true, "mapped" : 24630, "mappedWithJournal" : 49260 }, "metrics" : { "document" : { "deleted" : NumberLong(4398459), "inserted" : NumberLong(43661297), "returned" : NumberLong(249187499), "updated" : NumberLong(868225290) }, "getLastError" : { "wtime" : { "num" : 1332687, "totalMillis" : 154 }, "wtimeouts" : NumberLong(0) }, "operation" : { "fastmod" : NumberLong(846861873), "idhack" : NumberLong(1085704909), "scanAndOrder" : NumberLong(0) }, "queryExecutor" : { "scanned" : NumberLong(1095979673) }, "record" : { "moves" : NumberLong(1249911) }, "repl" : { "apply" : { "batches" : { "num" : 0, "totalMillis" : 0 }, "ops" : NumberLong(0) }, "buffer" : { "count" : NumberLong(0), "maxSizeBytes" : 268435456, "sizeBytes" : NumberLong(0) }, "network" : { "bytes" : NumberLong(0), "getmores" : { "num" : 0, "totalMillis" : 0 }, "ops" : NumberLong(0), "readersCreated" : NumberLong(0) }, "oplog" : { "insert" : { "num" : 0, "totalMillis" : 0 }, "insertBytes" : NumberLong(0) }, "preload" : { "docs" : { "num" : 0, "totalMillis" : 0 }, "indexes" : { "num" : 0, "totalMillis" : 0 } } }, "ttl" : { "deletedDocuments" : NumberLong(10), "passes" : NumberLong(25857) } }, "ok" : 1 }
锁确实比较多，查看mongotop
root@localhost:/mnt/mongodb/log# mongotop connected to: 127.0.0.1 ns total read write 2017-07-14T06:46:41 countly.online_users55d2e563292d6b2c7d1d132d 1669ms 1665ms 4ms countly.online_users5770b587f4a7f02a6a2b9cce 53ms 52ms 1ms countly.carriers 42ms 0ms 42ms countly.online_users5506ab1979a3a67e5214cda6 40ms 38ms 2ms countly.devices 16ms 0ms 16ms countly.apps 9ms 9ms 0ms countly.users 8ms 0ms 8ms countly.device_details 7ms 0ms 7ms countly.app_users55d2e563292d6b2c7d1d132d 4ms 0ms 4ms ns total read write 2017-07-14T06:46:42 countly.online_users55d2e563292d6b2c7d1d132d 3671ms 3669ms 2ms countly.online_users5770b587f4a7f02a6a2b9cce 64ms 63ms 1ms countly.online_users5506ab1979a3a67e5214cda6 26ms 25ms 1ms countly.devices 19ms 0ms 19ms countly.carriers 16ms 0ms 16ms countly.users 6ms 0ms 6ms countly.apps 6ms 6ms 0ms countly.app_users55d2e563292d6b2c7d1d132d 4ms 1ms 3ms countly_drill.drill_eventse02cd05a335f76036692df25f94be2b3d5c20bd2 3ms 0ms 3ms ns total read write 2017-07-14T06:46:43 countly.online_users55d2e563292d6b2c7d1d132d 1709ms 1705ms 4ms countly.online_users5506ab1979a3a67e5214cda6 33ms 31ms 2ms countly.carriers 19ms 0ms 19ms countly.devices 10ms 0ms 10ms countly.users 10ms 0ms 10ms countly.online_users54feb14879a3a67e52f6f7cd 7ms 7ms 0ms countly.apps 5ms 5ms 0ms countly.app_users55d2e563292d6b2c7d1d132d 4ms 0ms 4ms countly_drill.drill_eventsbb446ae67767dec699789c496e6c067b32dee999 3ms 0ms 3ms ns total read write 2017-07-14T06:46:44 countly.online_users55d2e563292d6b2c7d1d132d 2301ms 2298ms 3ms countly.online_users5770b587f4a7f02a6a2b9cce 262ms 261ms 1ms countly.online_users5506ab1979a3a67e5214cda6 35ms 34ms 1ms countly_drill.drill_eventse02cd05a335f76036692df25f94be2b3d5c20bd2 22ms 0ms 22ms countly.online_users54feb14879a3a67e52f6f7cd 8ms 7ms 1ms countly.users 7ms 0ms 7ms countly_drill.drill_eventscc860280b73e64c3413b38f2e9959acb7babd3f9 6ms 0ms 6ms countly.app_users55d2e563292d6b2c7d1d132d 5ms 2ms 3ms countly.apps 4ms 4ms 0ms
是读的时间很长，这个怀疑是没有建立索引，另外mongodb本身有一个bug
https://docs.mongodb.com/manual/administration/production-notes/
解决方法如下所示
#numactl --interleave all command
原理如下
http://www.cnblogs.com/Lifehacker/p/database_swap_insanity_on_Linux.html
这么操作了好像也没什么效果，最后想到之前这个数据库因为磁盘爆满清空过，之前听说过mongodb迁移后会丢失索引，要重建索引，我查了一下索引
> db.online_users55d2e563292d6b2c7d1d132d.find() { "_id" : "4e5d3922-6553-430a-849f-8b1007373d7c", "la" : ISODate("2017-06-30T00:55:08.933Z"), "n" : 0 } { "_id" : "55d7d514-f75c-432a-b557-17c036437a32", "la" : ISODate("2017-06-29T16:38:12.863Z"), "n" : 0 } { "_id" : "0118022e-54d2-4fb9-998d-a544d9a76d42", "la" : ISODate("2017-07-05T17:17:44.307Z"), "n" : 0 } { "_id" : "2a160945-fcc1-42f5-8fd5-693f97c4c332", "la" : ISODate("2017-07-02T07:01:35.619Z"), "n" : 0 } { "_id" : "14cd09d4-fd09-43b8-9259-8a007fc08b45", "la" : ISODate("2017-06-28T01:23:29.400Z"), "n" : 0 } { "_id" : "b9fa7cc9-cd8c-4a1b-83fb-fc4c4f929715", "la" : ISODate("2017-06-26T13:38:32.887Z"), "n" : 1 } { "_id" : "1d968b85-2571-4cf7-9520-bfeb44d893bc", "la" : ISODate("2017-07-13T13:08:18.807Z"), "n" : 0 } { "_id" : "0b1a3d07-8e5d-4e78-a93f-d13811a81b86", "la" : ISODate("2017-07-13T03:10:39.927Z"), "n" : 0 } { "_id" : "18d8f9fb-dff3-46e6-a3c7-bb19db748c4a", "la" : ISODate("2017-07-08T17:59:37.518Z"), "n" : 0 } { "_id" : "a81349ac-9d83-4ef7-be0b-1b9bf64f867f", "la" : ISODate("2017-07-13T23:27:35.174Z"), "n" : 0 } { "_id" : "dd841258-36a6-45de-bc45-450441d9bf33", "la" : ISODate("2017-07-14T06:34:23.838Z"), "n" : 0 } { "_id" : "75720c9f-69af-4833-ae14-78a381cc31b6", "la" : ISODate("2017-07-13T15:57:28.700Z"), "n" : 0 } { "_id" : "ef37ff3d-979d-4275-9991-398ccb2d5576", "la" : ISODate("2017-07-14T04:45:38.872Z"), "n" : 0 } { "_id" : "465f923e-b0f1-4aa7-8f68-2d7db95973b2", "la" : ISODate("2017-07-07T11:47:53.029Z"), "n" : 0 } { "_id" : "e6b5fd38-bb56-4634-9232-7eeceb8cf46a", "la" : ISODate("2017-07-09T03:27:04.545Z"), "n" : 0 } { "_id" : "d4957c5d-154f-4c4f-8843-9792beab4817", "la" : ISODate("2017-07-11T21:49:35.417Z"), "n" : 0 } { "_id" : "08d0d626-b37b-4761-81b2-98b5ee426df8", "la" : ISODate("2017-06-29T13:56:14.974Z"), "n" : 0 } { "_id" : "6d2eef45-cb25-4ff6-b8cd-497ab3b938ae", "la" : ISODate("2017-07-07T02:19:16.154Z"), "n" : 0 } { "_id" : "6f444a2b-e2fc-4ab0-8b1d-d4334bf095e3", "la" : ISODate("2017-07-09T12:17:05.711Z"), "n" : 0 } { "_id" : "85e06ac1-7158-42fb-95ed-2cd39a7f0687", "la" : ISODate("2017-06-26T02:27:25.389Z"), "n" : 1 } Type "it" for more > db.online_users55d2e563292d6b2c7d1d132d.find({"_id" : "4e5d3922-6553-430a-849f-8b1007373d7c"}).explain() { "cursor" : "BtreeCursor _id_", "isMultiKey" : false, "n" : 1, "nscannedObjects" : 1, "nscanned" : 1, "nscannedObjectsAllPlans" : 1, "nscannedAllPlans" : 1, "scanAndOrder" : false, "indexOnly" : false, "nYields" : 0, "nChunkSkips" : 0, "millis" : 0, "indexBounds" : { "start" : { "_id" : "4e5d3922-6553-430a-849f-8b1007373d7c" }, "end" : { "_id" : "4e5d3922-6553-430a-849f-8b1007373d7c" } }, "server" : "localhost:27017" }
明明是走索引的，可能是la这一列没有加索引（这个我当时没有看，但是从现在来看这个思路应该是正确的）
加上索引
db.online_users547c2a634dcc1eef3d000001.ensureIndex({la: 1}, {expireAfterSeconds: 180});
因为这个是cache性质频繁读写的表，要转成Capped Collection，自动老化移出，优化性能，并设置最大记录数
db.runCommand({"convertToCapped":"online_users547c2a634dcc1eef3d000001",size:60})
db.runCommand({"convertToCapped":"live_data547c2a634dcc1eef3d000001",size:60})
这样之后结果就是好多了，再观察一段时间吧