MongoDB for DBA's 2/7. Crud

Chapter 2. CRUD: Creating, reading and updating data

Insert operation

show dbs
use database
db --> show current database
show collections --> show the current database collections
db.sample.insert ( { a : 1 }) --> to insert a document
db.sample.find( )  --> to query documents
db.sample.find( ).pretty()  --> to query documents in compact format

a documents can have differents schemes.

var aa = db.sample.find( ).toArray()   --> put the result data in an array and assign to a variable aa like javascript

db.getLastError()  -->  to force the acknowlodge of the insert operation


Insert the document
{ x : 3 , y : 4 }
into the temperature collection for the current database. 

The collection will initially be empty, but you can (and should) check for your document after insertion. The shell will automatically add an _id field to the inserted document. You should only insert the one document.


> db.temperature.insert({x:3,y:4})
> WriteResult({ "nInserted" : 1 })
> db.temperature.find()
{ "_id" : ObjectId("54b5835f132c1f084b02ab21"), "y" : 4, "x" : 3 }

Update operation

you can update one o more documents in a collection. 
db.colection.update (<<where>>, 
                                 <<doc or partial update  expression in JSON or BSON format >> 
                                 , <upsert>  --> optional
                                 , <multi>  -->  optional
                                 )
There are two kinds of updates:

full document update replacement

partial update: only change a single field

<upsert> : update or insert if not present
<multi>  : many documents not only one
t = db.sample
t.update( { "_id" : 100}, { "a" : 100 } ) --> the id field must not change instead of we get an error

myobj = t.findOne()
{ "_id" : 190, x : "hello" }
myobj.y = 123

{ "_id" : 190, x : "hello", y : 123 }
t.update( { "_id" : myobj._id}, myobj )

Check all that are true about the _id field:




It must be of type ObjectId()It must exist in every documentIt can be changedIt must be unique inside the collectionIt is automatically indexed

Document growth / relocation


In v2.6 and higher, by default --> powerOf2sizes = on

Having powerOf2sizes on, the smallest documents you insert will be allocated in 32 byte allocation units. Then, it it's greater than that size, the next size will be 64 bytes, 128 bytes, etc. until to 16Mbytes limit. After 4MB it actually adds only 1MB at a time since we will getting closed to the upper limit .

The reason for this sizing strategy is to avoid heap fragmentation or deleted space fragmentation.
If you have a collection where you really needed to optimize, and you wanted a very specific allocation unit size you could turn this off.

What operations could cause an existing document to be moved on disk? Check all that apply.

Adding items to an arrayRemoving a fieldChanging a fields valueAdding new fields to an existing documentInserting a sub-document into an existing document

Save() command


It's a shell helper function, it's not a mongoDB server operation,

myobj = t.findOne()
{ "_id" : 190, x : "hello" }
myobj.y = 897

t.save(myobj)

{ "_id" : 190, x : "hello", y : 897 }

What happens if you try to use db.collection.save(document) if the inserted document has no _id?

It will assign the document a new _id of the type already used in the table and insert it.It will assign the document an existing _id field at random and overwrite that existing document.It will insert the document with no _id.It will infer on its own what you meant and do that.It will assign the document an objectID for its _id field, and then insert it.It will throw an exception, as it cannot update a document that does not exist in the collection.

Partial Updates & Documents

The size limit for each document is 16MBytes in order not to saturate the ethernet connection in  high documents like that 100MB.

We can use operators like:

1. $set : to set a new value
2. $push: to add somethig to an array. If it's not exist, it will be created
3. $addToSet: to add to an array if not already present
4. $pop : to remove from an array
5. $unset: to delete a field
6. etc ...

The Mongo Web Shell has been initialized with one collection, cars, with one document preexisting:
{ "_id" : 100, "name" : "GTO", "year" : 1969, "color" : "red" }
Set the available field to 1. 

You can verify the initial state of the collection, and verify your answer before you submit. There should be only one document in the collection. 

> db.cars.find()
{ "_id" : 100, "color" : "red", "name" : "GTO", "year" : 1969 }
> db.cars.update({ "_id":100}, { "$set":{"available":1}})
WriteResult({ "nMatched" : 1, "nUpserted" : 0, "nModified" : 1 })
> db.cars.find()
{
    "_id" : 100,
    "color" : "red",
    "available" : 1,
    "name" : "GTO",
    "year" : 1969
}


Removing Documents

db.<< collection name>>.remove( << expression >>)

This is a multidocument operation.

db.test.remove( { "_id" : 100 } )

db.test.remove( { "x" : /hello/ })  --> with regular expressions


We have initialized documents in the users collection of the form:

{
 _id : ObjectId("50897dbb9b96971d287202a9"),
 name : "Jane",
 likes : [ "tennis", "golf" ],
 registered : false,
 addr : {
   city : "Lyon",
   country : "France" 
 } 
}

Delete all documents in the collection where city is "Lyon" and registered is false. 

You can check the initial state of the collection, and verify your answer before you submit. 

> db.users.count()

300

> db.users.remove({ "addr.city" : "Lyon", "registered" : false})

WriteResult({ "nRemoved" : 33 })

> db.users.count()

267

Multi Update

db.collection.update( query_document , update_document , [ options_document ] )

where optional options_document has any one or more of the following optional parameters:

upsert : true/false,
multi : true/false,
writeConcern: document

by default --> upsert : false, multi : false

Which of the following are disadvantages to setting multi=false (as it is by default)?

Updates on non-indexed query fields will take longer.Updates on indexed query fields will take longer.Updates that the user may have intended to match multiple documents will exit prematurely, after only one update.Updates that match only one document will take longer.

Upsert

updates or inserts if not present


We have initialized documents in the users collection, one of which is the following:

{
 _id : "Jane",
 likes : [ "tennis", "golf" ],
 registered : false,
 addr : {
   city : "Lyon",
   country : "France" 
 } 
}

In the shell, add that this user likes "football". You should not need to pass "tennis" or "golf" to theupdate query at all. 

You can look at the initial state of the collection, and verify your answer before you submit. 

> db.users.count()

11

> db.users.find( { "likes" : "football" })

> db.users.update( {}, { "$addToSet" :{ "likes" : "football" }}, { "upsert":true,"multi":true})

WriteResult({ "nMatched" : 11, "nUpserted" : 0, "nModified" : 11 })

Wire Protocol

What are the basic building blocks of the wire protocol? Check all that apply.

GetMoreInsertQueryRemoveUpdateCommand

Which operation is overloaded in the wire protocol to handle commands?

GetMoreRemoveQueryInsertUpdate

Bulk() write Operations & Methods

There are two basic forms:

1. ordered: With an ordered list of operations, MongoDB executes the operations serially. If an error occurs during the processing of one of the write operations, MongoDB will return without processing any remaining write operations in the list.
2. unordered: With an unordered list of operations, MongoDB can execute the operations in parallel. If an error occurs during the processing of one of the write operations, MongoDB will continue to process remaining write operations in the list.

To use the Bulk() methods:

1. Initialize a list of operations using either db.collection.initializeUnorderedBulkOp() ordb.collection.initializeOrderedBulkOp().
2. Add write operations to the list using the following methods:
• Bulk.insert()
• Bulk.find()
• Bulk.find.upsert()
• Bulk.find.update()
• Bulk.find.updateOne()
• Bulk.find.replaceOne()
• Bulk.find.remove()
• Bulk.find.removeOne()
3. To execute the list of operations, use the Bulk.execute() method. You can specify the write concern for the list in the Bulk.execute() method.

Once executed, you cannot re-execute the list without reinitializing.

For example,

var bulk = db.items.initializeUnorderedBulkOp();

bulk.insert( { _id: 1, item: "abc123", status: "A", soldQty: 5000 } );

bulk.insert( { _id: 2, item: "abc456", status: "A", soldQty: 150 } );

bulk.insert( { _id: 3, item: "abc789", status: "P", soldQty: 0 } );

bulk.execute( { w: "majority", wtimeout: 5000 } );

Common Commands

1. user commands
1. getLastError()
2. isMaster()
3. aggregation operations
4. map-reduce
5. count
6. findAndModify: for complex updates. Nice way to do atomic, like, dequeuing of things from a queue
2. dba commands
1. drop collection
2. create collection
3. compact collection
4. serverStatus
5. replSetGetStatus()
6. addShard

Which of the following are user commands (as opposed to admin commands)? Check all that apply.

countaggregatedropMapReduceaddShardFindAndModifyserverStatus

db.runCommand()

db.runCommand(  {  << command name >> : <value>, param1: value1, param2 : value2} )

to get confirmations when the last operation was completed:

db.runCommand(  { "getLastError" : 1,
                                 "w"  : 2
                                 "wtimeout" : 3 }
                           )
When we execute an operation through the shell or another interface, we can use which of the following?

db.teaEarlGreyHot()db.engage()db.runCommand()db.makeItSo()

db.isMaster()

db.runCommand( "isMaster" )
db.isMaster()
db.runCommand( { "isMaster" : 1 } )  --> tell us if the server we are talking to is primary or not

What does the db.isMaster() command do?

Determine your level of mastery of the MongoDB shell.Determine if this shard is the Primary shard.Determine if this member of the replica set is the Primary.Determine if the database has Master-Slave replication.

db.serverStatus()

db.serverStatus() --> gives a lot of the statistics on what is happening on the server and a alot of those are used by the  MMS  (Mongo Monitoring Service)

When running the command db.serverStatus() in the shell, what does the “ok” field represent?

That the command was sent and received properly by the mongodThat the master db process is okThat the shell is running properlyThat the daemon to which you are connected is okThat the daemon has master status

db.currentOp() & db.killOp()

db.currentOp() --> shows information of the current operation
db.killOp() --> to stop a long-running operation

If you’re looking for problems with database performance, what is a good place to look, when you run db.currentOp()?

“ok”: false“secs_running”: (look for long times)“secs_running”: (look for short times)"opid", because you always want to killOp() everything to clean things up."opid": (something that is not sequential with the rest)

ensureIndex(), getIndexes() & dropIndex()

db.collection.ensureIndex( { "name" : 1} ) --> to create a index
db.collection.getIndexes() --> get the indexes of a collection
db.collection.dropIndex(index) --> drop th indexes of a collection

db.products.find().explain() --> give us information about how the find was done

What will happen if an index is created on a field that does not exist in any of the documents in the collection?

MongoDB will throw an error and not create the indexMongoDB will not create the index and not throw an errorMongoDB will create the index but give a warningMongoDB will create the index but give no feedbackMongoDB will ask if you're sure first

collection.stats() & collection.drop()

db.collection.stats() --> give us statistical information about the collection
db.collection.drop() --> remove collection even though from namespaces

True or false: db.collection.remove({}), which removes all messages in a collection, is the same as db.collection.drop(), which drops the collection.

TrueFalse

Review of Commands

Commands:

1. Server
1. isMaster
2. serverStatus
3. logout
4. getLastError
2. db
1. dropDatabase
2. repairDatabase
3. clone
4. copydb
5. dbStats
3. collection
1. DBA
1. create --> is implicit
2. drop
3. collstats
4. rename collection
2. user
1. count
2. aggregate
3. MapReduce
4. findAndModify
5. geo*
4. index
1. ensureIndex
2. getIndex

Which of these statements is true?

MongoDB commands are for administrator use only.In MongoDB collections are usually created implicitly.In MongoDB indexes are usually created implicitly.