One .mongodb Monitoring of

mongodb Can pass profile To monitor data , To optimize .

See if it's on profile Function with command

db.getProfilingLevel()   return level Grade , The value is 0|1|2, Respectively means ：0 On behalf of closed ,1 For recording slow orders ,2 Represent all

Start profile The function is

db.setProfilingLevel(level);  #level Grade , The value is the same as above.

level by 1 When , The default value of slow command is 100ms, Change to db.setProfilingLevel(level,slowms) Such as db.setProfilingLevel(1,50) This changes to 50 millisecond

adopt db.system.profile.find() View the current monitoring log .

Such as ：

  1. > db.system.profile.find({millis:{$gt:500}})
  2. { "ts" : ISODate("2011-07-23T02:50:13.941Z"), "info" : "query order.order reslen:11022 nscanned:672230  \nquery: { status: 1.0 } nreturned:101 bytes:11006 640ms", "millis" : 640 }
  3. { "ts" : ISODate("2011-07-23T02:51:00.096Z"), "info" : "query order.order reslen:11146 nscanned:672302  \nquery: { status: 1.0, user.uid: { $gt: 1663199.0 } }  nreturned:101 bytes:11130 647ms", "millis" : 647 }

The value here means

ts： Command execution time

info： The content of the order

query： Representative query

order.order： Represents the library and collection of queries

reslen： Returned result set size ,byte Count

nscanned： Number of scan records

nquery： After that is the query criteria

nreturned： Return the number of records and time used

millis： Time spent

If it takes a long time , Then we need to optimize .

such as nscanned It's a big number , Or close to the total number of records , Then you may not use index queries .

reslen It's big , It is possible to return unnecessary fields .

nreturned It's big , Then it is possible to query without restrictions .

mongo Can pass db.serverStatus() see mongod Operating state

  1. > db.serverStatus()
  2. {
  3. "host" : "baobao-laptop",# Host name
  4. "version" : "1.8.2",# Version number
  5. "process" : "mongod",# Process name
  6. "uptime" : 15549,# The elapsed time
  7. "uptimeEstimate" : 15351,
  8. "localTime" : ISODate("2011-07-23T06:07:31.220Z"), current time
  9. "globalLock" : {
  10. "totalTime" : 15548525410,# Total operation time （ns）
  11. "lockTime" : 89206633,  # Total lock time （ns）
  12. "ratio" : 0.005737305027178137,# Lock ratio
  13. "currentQueue" : {
  14. "total" : 0,# The current queue to execute
  15. "readers" : 0,# Read the queue
  16. "writers" : 0# Write a queue
  17. },
  18. "activeClients" : {
  19. "total" : 0,# The number of links executed by the current client
  20. "readers" : 0,# Number of read Links
  21. "writers" : 0# Write Links
  22. }
  23. },
  24. "mem" : {# The memory of
  25. "bits" : 32,#32 Bit system
  26. "resident" : 337,# The amount of physical memory occupied
  27. "virtual" : 599,# Occupy virtual memory
  28. "supported" : true,# Whether it supports expanding memory
  29. "mapped" : 512
  30. },
  31. "connections" : {
  32. "current" : 2,# Current number of links
  33. "available" : 817# Number of links available
  34. },
  35. "extra_info" : {
  36. "note" : "fields vary by platform",
  37. "heap_usage_bytes" : 159008,# Heap usage bytes
  38. "page_faults" : 907 # The page pretends to be
  39. },
  40. "indexCounters" : {
  41. "btree" : {
  42. "accesses" : 59963, # The number of indexes accessed
  43. "hits" : 59963, # So the number of hits
  44. "misses" : 0,# Index deviation number
  45. "resets" : 0,# Reset number
  46. "missRatio" : 0# Miss rate
  47. }
  48. },
  49. "backgroundFlushing" : {
  50. "flushes" : 259,  # Refresh times
  51. "total_ms" : 3395, # Total refresh time
  52. "average_ms" : 13.108108108108109, # Average duration
  53. "last_ms" : 1, # The last time
  54. "last_finished" : ISODate("2011-07-23T06:07:22.725Z")# Last refresh time
  55. },
  56. "cursors" : {
  57. "totalOpen" : 0,# Number of open cursors
  58. "clientCursors_size" : 0,# Client cursor size
  59. "timedOut" : 16# Timeout time
  60. },
  61. "network" : {
  62. "bytesIn" : 285676177,# input data （byte）
  63. "bytesOut" : 286564,# Output data （byte）
  64. "numRequests" : 2012348# Number of requests
  65. },
  66. "opcounters" : {
  67. "insert" : 2010000, # Insert operands
  68. "query" : 51,# Query operands
  69. "update" : 5,# Update operands
  70. "delete" : 0,# Delete operands
  71. "getmore" : 0,# Get more operands
  72. "command" : 148# Other command operands
  73. },
  74. "asserts" : {# The number of assertions
  75. "regular" : 0,
  76. "warning" : 0,
  77. "msg" : 0,
  78. "user" : 2131,
  79. "rollovers" : 0
  80. },
  81. "writeBacksQueued" : false,
  82. "ok" : 1
  83. }

db.stats() View the original status of a library

  1. > db.stats()
  2. {
  3. "db" : "order",# Library name
  4. "collections" : 4,# Set number
  5. "objects" : 2011622,# Record number
  6. "avgObjSize" : 111.92214441878245,# The average of each record
  7. "dataSize" : 225145048,# The total size of the record
  8. "storageSize" : 307323392,# Pre allocated storage space
  9. "numExtents" : 21,# Number of events
  10. "indexes" : 1,# Index number
  11. "indexSize" : 74187744,# So the size
  12. "fileSize" : 1056702464,# file size
  13. "ok" : 1
  14. }

To view the set record

  1. > db.order.stats()
  2. {
  3. "ns" : "order.order",# Namespace
  4. "count" : 2010000,# Record number
  5. "size" : 225039600,# size
  6. "avgObjSize" : 111.96,
  7. "storageSize" : 307186944,
  8. "numExtents" : 18,
  9. "nindexes" : 1,
  10. "lastExtentSize" : 56089856,
  11. "paddingFactor" : 1,
  12. "flags" : 1,
  13. "totalIndexSize" : 74187744,
  14. "indexSizes" : {
  15. "_id_" : 74187744# The index for _id_ The index size of
  16. },
  17. "ok" : 1
  18. }

mongostat Command to view running real-time statistics , Represents the number of real-time execution per second

mongodb It also provides an opportunity http Monitoring page of , You can visit http://ip:28017 Check it out. , This page is basically a synthesis of the above commands , So I won't go into details here .

Two .mongodb The optimization of the

According to the above monitoring methods , After finding the problem , We can optimize

There's a slow command on it , Now we can track it by executing the plan , Such as

  1. > db.order.find({ "status": 1.0, "user.uid": { $gt: 2663199.0 } }).explain()
  2. {
  3. "cursor" : "BasicCursor",# cursor type
  4. "nscanned" : 2010000,# Number of scans
  5. "nscannedObjects" : 2010000,# Scanning objects
  6. "n" : 337800,# Return the data
  7. "millis" : 2838,# Time consuming
  8. "nYields" : 0,
  9. "nChunkSkips" : 0,
  10. "isMultiKey" : false,
  11. "indexOnly" : false,
  12. "indexBounds" : {# Use index （ Nothing here ）
  13. }
  14. }

For something like this , We can create indexes

Can pass  db.collection.ensureIndex({" Field name ":1}) To create an index ,1 In ascending order ,-1 For the descending order , In the case of multiple data already available , It can be executed in the background , sentence db.collection.ensureIndex({" Field name ":1} , {backgroud:true})

Get the reference db.collection.getIndexes() see

Here we create a user.uid The index of >db.order.ensureIndex({"user.uid":1})

Re execute after creation

  1. db.order.find({ "status": 1.0, "user.uid": { $gt: 2663199.0 } }).explain()
  2. {
  3. "cursor" : "BtreeCursor user.uid_1",
  4. "nscanned" : 337800,
  5. "nscannedObjects" : 337800,
  6. "n" : 337800,
  7. "millis" : 1371,
  8. "nYields" : 0,
  9. "nChunkSkips" : 0,
  10. "isMultiKey" : false,
  11. "indexOnly" : false,
  12. "indexBounds" : {
  13. "user.uid" : [
  14. [
  15. 2663199,
  16. 1.7976931348623157e+308
  17. ]
  18. ]
  19. }
  20. }

Fewer scans , Speed up .mongodb The design of index is similar to that of relational database , Search by index to speed up reading , But too much pressure on writing , So I'm not going to talk about it here .

2. Other optimizations can use hint Force index lookup , Return just the data you need , Paging data, etc .

mongodb More articles on monitoring and performance optimization of

  1. Dynamic CRM 2013 Learning notes （ 7、 ... and ） track 、 Monitoring and performance optimization

    This article will introduce CRM Three content tracking of . Monitoring and performance optimization . Tracking is CRM It's a very useful function in , It can serve our CRM Debug or fix errors . Warnings provide valuable information : We can use window Performance monitoring tools to understand CRM The performance status of : Last ...

  2. Oracle Management monitoring section space utilization monitoring -oracle performance optimization

    SELECT S.OWNER,       S.SEGMENT_NAME,       S.SEGMENT_TYPE,       S.PARTITION_NAME,       ROUND(BYTE ...

  3. Mongodb Performance optimization problem of

    Abstract Database performance has a crucial impact on the overall performance of software , about Mongodb The common performance optimization methods of database mainly include : Paradigmatization and anti paradigmatization : The use of fill factor : Use of index : One . Paradigmatization and anti paradigmatization Paradigms are about eliminating duplicate data ...

  4. iOS Performance optimization summary

    iOS Performance optimization summary . About iOS Performance optimization : Basic tools . Business optimization . Memory optimization . Carton optimization . Layout optimization . Power optimization . Install the bag to slim down . Startup optimization . Network optimization, etc . About iOS Performance optimization : Basic tools . Business optimization . Memory ...

  5. mongodb Can pass profile To monitor data (mongodb performance optimization )

    mongodb Can pass profile To monitor data (mongodb performance optimization )   Turn on Profiling   function  , Optimize slow queries : mongodb Can pass profile To monitor data , To optimize . see ...

  6. MongoDB performance optimization

    One . Indexes MongoDB Provides diverse indexing support , The index information is stored in system.indexes in , And the default is always _id Create index , Its index uses basic and MySQL Like a relational database . In fact, it can be said that , Indexes ...

  7. MongoDB Performance optimization guidelines

    One . Indexes MongoDB Provides diverse indexing support , The index information is stored in system.indexes in , And the default is always _id Create index , Its index uses basic and MySQL Like a relational database . In fact, it can be said that , Indexes ...

  8. Mongodb Advanced - performance optimization

    1. monitor mongodb Can pass profile To monitor data , To optimize . See if it's on profile Function with command :db.getProfilingLevel() return level Grade , The value is 0|1|2, Represent the ...

  9. MongoDB Learning notes （ Four ）-- Indexes && performance optimization

    Indexes                                                                                              Basic index ...

Random recommendation

  1. Swipecards

    https://github.com/Diolor/Swipecards https://github.com/kikoso/Swipeable-Cards

  2. ios Project development summary

    UI Interface iOS and Android Interface design dimension specification  http://www.alibuybuy.com/posts/85486.html iPhone app Interface design dimension specification  http://www. ...

  3. Explain profound theories in simple language ES6： Indefinite parameters and default parameters

    Uncertain parameters We usually use variable parameter functions to construct API, Variable parameter functions can take any number of parameters . for example ,String.prototype.concat Method can accept any number of string parameters .ES6 Provides a new way to write variable parameter functions ...

  4. Alibaba cloud uses js Realization OSS Image upload 、 obtain OSS Picture list 、 Get the picture Internet access address ( Read write access is private 、 Read and write access public );

    Please refer to :https://help.aliyun.com/document_detail/32069.html?spm=a2c4g.11186623.6.763.ZgC59a perhaps https://h ...

  5. 【Qt】 The window is centered

    w.move((a.desktop()->width() - w.width())/, (a.desktop()->height() - w.height())/); The above method can be put in the middle , but ...

  6. 10-03 Java Overview and explanation of the package

    Compile and run with package A: Manual a: Write a java file . b: adopt javac Command to compile the java file . c: Manually create the package name . d: hold b The steps of the class Files in c The lowest level package of the step e: Go back to the same location as the package root ...

  7. sysfs File system learning --sysfs

    One .sysfs brief introduction 1.sysfs Is the use VFS Interface to read and write kobject Hierarchical structure , File systems built up . The updates and deletions are those xxx_register()/unregister() What to do love . from sysfs ...

  8. JavaScript in Function Declaration And Function Expression Or say function fn(){} and var fn=function(){} The difference between

    JavaScript It's an interpretative language , The function declaration will be in JavaScript After the code loads . Explained before execution , Function expressions are interpreted only when executed to this line of code . stay JS There are two ways to define functions in , 1 yes :var aaa= ...

  9. 160705、 summary ：commons-codec.jar Common methods in

    One .Base64 Encoding and decoding import org.apache.commons.codec.EncoderException;import org.apache.commons.codec.binary ...

  10. flask The signal

    Sao Shi blog : The signal That's how you understand the signal , Request is compared to racing car , The process of asking to go is racing track , And the signal is located in the gas station and maintenance station on the racetrack , The function of signal registration is like the person in the maintenance station , Every time I pass the maintenance station and someone in the maintenance station carries out maintenance The signal is understood here : ...