我在解析表以创建视图的表中有一个JSON列。我在JSON字符串中有120列，它产生大约1750行数据。问题是，查询需要17秒钟才能运行，这是一个基本查询，类似于下面的查询（已选择120列）：
SELECT JSON_VALUE(json, '$.member.firstname') as MemberFirstName
, JSON_VALUE(json, '$.member.lastname') as MemberLastName
, JSON_VALUE(json, '$.member.phone') as MemberPhone
FROM foobar
我知道该表将不断增长，并且我希望在使用代码创建视图之前提高其性能。我正在阅读一些有关创建计算列或使用DBCC PAGE函数的文章，但我想知道是否有一种强度较低的方法来获得我们想要的东西。
答案 0 :(得分：4)
根据上面的查询，我对json的结构做了一些假设。
OPENJSON在JSON文档上提供了一个行集视图，基本上将其视为一个表，而JSON_VALUE（）从JSON字符串中提取一个值，这意味着它将在每一行的每一列上执行该操作，并且会更加昂贵。
这里是一个示例，因此您可以看到使用JSON_VALUE()与OPENJSON()的区别。 OPENJSON（）将需要更大的查询，但整体效果会更好。
--Setup some test data so we can see performance with a similar number of columns you're dealing with.
CREATE TABLE [#foobar]
(
[jsondata] NVARCHAR(MAX)
);
--load our temp table with 2000 test records
DECLARE @Counter INT = 1;
WHILE @Counter <= 2000
BEGIN
INSERT INTO [#foobar] (
[jsondata]
)
VALUES (
N'{
"member": {
"firstname": "John",
"lastname": "Smith",
"phone": "1234566789",
"column1": "testdata1"
,"column2": "testdata2"
,"column3": "testdata3"
,"column4": "testdata4"
,"column5": "testdata5"
,"column6": "testdata6"
,"column7": "testdata7"
,"column8": "testdata8"
,"column9": "testdata9"
,"column10": "testdata10"
,"column11": "testdata11"
,"column12": "testdata12"
,"column13": "testdata13"
,"column14": "testdata14"
,"column15": "testdata15"
,"column16": "testdata16"
,"column17": "testdata17"
,"column18": "testdata18"
,"column19": "testdata19"
,"column20": "testdata20"
,"column21": "testdata21"
,"column22": "testdata22"
,"column23": "testdata23"
,"column24": "testdata24"
,"column25": "testdata25"
,"column26": "testdata26"
,"column27": "testdata27"
,"column28": "testdata28"
,"column29": "testdata29"
,"column30": "testdata30"
,"column31": "testdata31"
,"column32": "testdata32"
,"column33": "testdata33"
,"column34": "testdata34"
,"column35": "testdata35"
,"column36": "testdata36"
,"column37": "testdata37"
,"column38": "testdata38"
,"column39": "testdata39"
,"column40": "testdata40"
,"column41": "testdata41"
,"column42": "testdata42"
,"column43": "testdata43"
,"column44": "testdata44"
,"column45": "testdata45"
,"column46": "testdata46"
,"column47": "testdata47"
,"column48": "testdata48"
,"column49": "testdata49"
,"column50": "testdata50"
,"column51": "testdata51"
,"column52": "testdata52"
,"column53": "testdata53"
,"column54": "testdata54"
,"column55": "testdata55"
,"column56": "testdata56"
,"column57": "testdata57"
,"column58": "testdata58"
,"column59": "testdata59"
,"column60": "testdata60"
,"column61": "testdata61"
,"column62": "testdata62"
,"column63": "testdata63"
,"column64": "testdata64"
,"column65": "testdata65"
,"column66": "testdata66"
,"column67": "testdata67"
,"column68": "testdata68"
,"column69": "testdata69"
,"column70": "testdata70"
,"column71": "testdata71"
,"column72": "testdata72"
,"column73": "testdata73"
,"column74": "testdata74"
,"column75": "testdata75"
,"column76": "testdata76"
,"column77": "testdata77"
,"column78": "testdata78"
,"column79": "testdata79"
,"column80": "testdata80"
,"column81": "testdata81"
,"column82": "testdata82"
,"column83": "testdata83"
,"column84": "testdata84"
,"column85": "testdata85"
,"column86": "testdata86"
,"column87": "testdata87"
,"column88": "testdata88"
,"column89": "testdata89"
,"column90": "testdata90"
,"column91": "testdata91"
,"column92": "testdata92"
,"column93": "testdata93"
,"column94": "testdata94"
,"column95": "testdata95"
,"column96": "testdata96"
,"column97": "testdata97"
,"column98": "testdata98"
,"column99": "testdata99"
,"column100": "testdata100"
,"column101": "testdata101"
,"column102": "testdata102"
,"column103": "testdata103"
,"column104": "testdata104"
,"column105": "testdata105"
,"column106": "testdata106"
,"column107": "testdata107"
,"column108": "testdata108"
,"column109": "testdata109"
,"column110": "testdata110"
,"column111": "testdata111"
,"column112": "testdata112"
,"column113": "testdata113"
,"column114": "testdata114"
,"column115": "testdata115"
,"column116": "testdata116"
,"column117": "testdata117"
,"column118": "testdata118"
,"column119": "testdata119"
,"column120": "testdata120"
}
}'
);
SET @Counter = @Counter + 1;
END;
使用JSON_VALUE（）运行原始查询，我在13到15秒的范围内：
SELECT JSON_VALUE([jsondata], '$.member.firstname') AS [MemberFirstName]
, JSON_VALUE([jsondata], '$.member.lastname') AS [MemberLastName]
, JSON_VALUE([jsondata], '$.member.phone') AS [MemberPhone]
, JSON_VALUE([jsondata], '$.member.column1') AS [Column1]
, JSON_VALUE([jsondata], '$.member.column2') AS [Column2]
, JSON_VALUE([jsondata], '$.member.column3') AS [Column3]
, JSON_VALUE([jsondata], '$.member.column4') AS [Column4]
, JSON_VALUE([jsondata], '$.member.column5') AS [Column5]
, JSON_VALUE([jsondata], '$.member.column6') AS [Column6]
, JSON_VALUE([jsondata], '$.member.column7') AS [Column7]
, JSON_VALUE([jsondata], '$.member.column8') AS [Column8]
, JSON_VALUE([jsondata], '$.member.column9') AS [Column9]
, JSON_VALUE([jsondata], '$.member.column10') AS [Column10]
, JSON_VALUE([jsondata], '$.member.column11') AS [Column11]
, JSON_VALUE([jsondata], '$.member.column12') AS [Column12]
, JSON_VALUE([jsondata], '$.member.column13') AS [Column13]
, JSON_VALUE([jsondata], '$.member.column14') AS [Column14]
, JSON_VALUE([jsondata], '$.member.column15') AS [Column15]
, JSON_VALUE([jsondata], '$.member.column16') AS [Column16]
, JSON_VALUE([jsondata], '$.member.column17') AS [Column17]
, JSON_VALUE([jsondata], '$.member.column18') AS [Column18]
, JSON_VALUE([jsondata], '$.member.column19') AS [Column19]
, JSON_VALUE([jsondata], '$.member.column20') AS [Column20]
, JSON_VALUE([jsondata], '$.member.column21') AS [Column21]
, JSON_VALUE([jsondata], '$.member.column22') AS [Column22]
, JSON_VALUE([jsondata], '$.member.column23') AS [Column23]
, JSON_VALUE([jsondata], '$.member.column24') AS [Column24]
, JSON_VALUE([jsondata], '$.member.column25') AS [Column25]
, JSON_VALUE([jsondata], '$.member.column26') AS [Column26]
, JSON_VALUE([jsondata], '$.member.column27') AS [Column27]
, JSON_VALUE([jsondata], '$.member.column28') AS [Column28]
, JSON_VALUE([jsondata], '$.member.column29') AS [Column29]
, JSON_VALUE([jsondata], '$.member.column30') AS [Column30]
, JSON_VALUE([jsondata], '$.member.column31') AS [Column31]
, JSON_VALUE([jsondata], '$.member.column32') AS [Column32]
, JSON_VALUE([jsondata], '$.member.column33') AS [Column33]
, JSON_VALUE([jsondata], '$.member.column34') AS [Column34]
, JSON_VALUE([jsondata], '$.member.column35') AS [Column35]
, JSON_VALUE([jsondata], '$.member.column36') AS [Column36]
, JSON_VALUE([jsondata], '$.member.column37') AS [Column37]
, JSON_VALUE([jsondata], '$.member.column38') AS [Column38]
, JSON_VALUE([jsondata], '$.member.column39') AS [Column39]
, JSON_VALUE([jsondata], '$.member.column40') AS [Column40]
, JSON_VALUE([jsondata], '$.member.column41') AS [Column41]
, JSON_VALUE([jsondata], '$.member.column42') AS [Column42]
, JSON_VALUE([jsondata], '$.member.column43') AS [Column43]
, JSON_VALUE([jsondata], '$.member.column44') AS [Column44]
, JSON_VALUE([jsondata], '$.member.column45') AS [Column45]
, JSON_VALUE([jsondata], '$.member.column46') AS [Column46]
, JSON_VALUE([jsondata], '$.member.column47') AS [Column47]
, JSON_VALUE([jsondata], '$.member.column48') AS [Column48]
, JSON_VALUE([jsondata], '$.member.column49') AS [Column49]
, JSON_VALUE([jsondata], '$.member.column50') AS [Column50]
, JSON_VALUE([jsondata], '$.member.column51') AS [Column51]
, JSON_VALUE([jsondata], '$.member.column52') AS [Column52]
, JSON_VALUE([jsondata], '$.member.column53') AS [Column53]
, JSON_VALUE([jsondata], '$.member.column54') AS [Column54]
, JSON_VALUE([jsondata], '$.member.column55') AS [Column55]
, JSON_VALUE([jsondata], '$.member.column56') AS [Column56]
, JSON_VALUE([jsondata], '$.member.column57') AS [Column57]
, JSON_VALUE([jsondata], '$.member.column58') AS [Column58]
, JSON_VALUE([jsondata], '$.member.column59') AS [Column59]
, JSON_VALUE([jsondata], '$.member.column60') AS [Column60]
, JSON_VALUE([jsondata], '$.member.column61') AS [Column61]
, JSON_VALUE([jsondata], '$.member.column62') AS [Column62]
, JSON_VALUE([jsondata], '$.member.column63') AS [Column63]
, JSON_VALUE([jsondata], '$.member.column64') AS [Column64]
, JSON_VALUE([jsondata], '$.member.column65') AS [Column65]
, JSON_VALUE([jsondata], '$.member.column66') AS [Column66]
, JSON_VALUE([jsondata], '$.member.column67') AS [Column67]
, JSON_VALUE([jsondata], '$.member.column68') AS [Column68]
, JSON_VALUE([jsondata], '$.member.column69') AS [Column69]
, JSON_VALUE([jsondata], '$.member.column70') AS [Column70]
, JSON_VALUE([jsondata], '$.member.column71') AS [Column71]
, JSON_VALUE([jsondata], '$.member.column72') AS [Column72]
, JSON_VALUE([jsondata], '$.member.column73') AS [Column73]
, JSON_VALUE([jsondata], '$.member.column74') AS [Column74]
, JSON_VALUE([jsondata], '$.member.column75') AS [Column75]
, JSON_VALUE([jsondata], '$.member.column76') AS [Column76]
, JSON_VALUE([jsondata], '$.member.column77') AS [Column77]
, JSON_VALUE([jsondata], '$.member.column78') AS [Column78]
, JSON_VALUE([jsondata], '$.member.column79') AS [Column79]
, JSON_VALUE([jsondata], '$.member.column80') AS [Column80]
, JSON_VALUE([jsondata], '$.member.column81') AS [Column81]
, JSON_VALUE([jsondata], '$.member.column82') AS [Column82]
, JSON_VALUE([jsondata], '$.member.column83') AS [Column83]
, JSON_VALUE([jsondata], '$.member.column84') AS [Column84]
, JSON_VALUE([jsondata], '$.member.column85') AS [Column85]
, JSON_VALUE([jsondata], '$.member.column86') AS [Column86]
, JSON_VALUE([jsondata], '$.member.column87') AS [Column87]
, JSON_VALUE([jsondata], '$.member.column88') AS [Column88]
, JSON_VALUE([jsondata], '$.member.column89') AS [Column89]
, JSON_VALUE([jsondata], '$.member.column90') AS [Column90]
, JSON_VALUE([jsondata], '$.member.column91') AS [Column91]
, JSON_VALUE([jsondata], '$.member.column92') AS [Column92]
, JSON_VALUE([jsondata], '$.member.column93') AS [Column93]
, JSON_VALUE([jsondata], '$.member.column94') AS [Column94]
, JSON_VALUE([jsondata], '$.member.column95') AS [Column95]
, JSON_VALUE([jsondata], '$.member.column96') AS [Column96]
, JSON_VALUE([jsondata], '$.member.column97') AS [Column97]
, JSON_VALUE([jsondata], '$.member.column98') AS [Column98]
, JSON_VALUE([jsondata], '$.member.column99') AS [Column99]
, JSON_VALUE([jsondata], '$.member.column100') AS [Column100]
, JSON_VALUE([jsondata], '$.member.column101') AS [Column101]
, JSON_VALUE([jsondata], '$.member.column102') AS [Column102]
, JSON_VALUE([jsondata], '$.member.column103') AS [Column103]
, JSON_VALUE([jsondata], '$.member.column104') AS [Column104]
, JSON_VALUE([jsondata], '$.member.column105') AS [Column105]
, JSON_VALUE([jsondata], '$.member.column106') AS [Column106]
, JSON_VALUE([jsondata], '$.member.column107') AS [Column107]
, JSON_VALUE([jsondata], '$.member.column108') AS [Column108]
, JSON_VALUE([jsondata], '$.member.column109') AS [Column109]
, JSON_VALUE([jsondata], '$.member.column110') AS [Column110]
, JSON_VALUE([jsondata], '$.member.column111') AS [Column111]
, JSON_VALUE([jsondata], '$.member.column112') AS [Column112]
, JSON_VALUE([jsondata], '$.member.column113') AS [Column113]
, JSON_VALUE([jsondata], '$.member.column114') AS [Column114]
, JSON_VALUE([jsondata], '$.member.column115') AS [Column115]
, JSON_VALUE([jsondata], '$.member.column116') AS [Column116]
, JSON_VALUE([jsondata], '$.member.column117') AS [Column117]
, JSON_VALUE([jsondata], '$.member.column118') AS [Column118]
, JSON_VALUE([jsondata], '$.member.column119') AS [Column119]
, JSON_VALUE([jsondata], '$.member.column120') AS [Column120]
FROM [#foobar];
这是一个使用WITH子句重写为OPENJSON（）的示例。这始终能在2-3秒内为我提供结果。
--Test using OPENJSON()
--You will want to list each column in your select from the with clause
--a lot better peformance then doing a select *
SELECT [b].[MemberFirstName]
, [b].[MemberLastName]
, [b].[MemberPhone]
, [b].[Column1]
, [b].[Column2]
, [b].[Column3]
, [b].[Column4]
, [b].[Column5]
, [b].[Column6]
, [b].[Column7]
, [b].[Column8]
, [b].[Column9]
, [b].[Column10]
, [b].[Column11]
, [b].[Column12]
, [b].[Column13]
, [b].[Column14]
, [b].[Column15]
, [b].[Column16]
, [b].[Column17]
, [b].[Column18]
, [b].[Column19]
, [b].[Column20]
, [b].[Column21]
, [b].[Column22]
, [b].[Column23]
, [b].[Column24]
, [b].[Column25]
, [b].[Column26]
, [b].[Column27]
, [b].[Column28]
, [b].[Column29]
, [b].[Column30]
, [b].[Column31]
, [b].[Column32]
, [b].[Column33]
, [b].[Column34]
, [b].[Column35]
, [b].[Column36]
, [b].[Column37]
, [b].[Column38]
, [b].[Column39]
, [b].[Column40]
, [b].[Column41]
, [b].[Column42]
, [b].[Column43]
, [b].[Column44]
, [b].[Column45]
, [b].[Column46]
, [b].[Column47]
, [b].[Column48]
, [b].[Column49]
, [b].[Column50]
, [b].[Column51]
, [b].[Column52]
, [b].[Column53]
, [b].[Column54]
, [b].[Column55]
, [b].[Column56]
, [b].[Column57]
, [b].[Column58]
, [b].[Column59]
, [b].[Column60]
, [b].[Column61]
, [b].[Column62]
, [b].[Column63]
, [b].[Column64]
, [b].[Column65]
, [b].[Column66]
, [b].[Column67]
, [b].[Column68]
, [b].[Column69]
, [b].[Column70]
, [b].[Column71]
, [b].[Column72]
, [b].[Column73]
, [b].[Column74]
, [b].[Column75]
, [b].[Column76]
, [b].[Column77]
, [b].[Column78]
, [b].[Column79]
, [b].[Column80]
, [b].[Column81]
, [b].[Column82]
, [b].[Column83]
, [b].[Column84]
, [b].[Column85]
, [b].[Column86]
, [b].[Column87]
, [b].[Column88]
, [b].[Column89]
, [b].[Column90]
, [b].[Column91]
, [b].[Column92]
, [b].[Column93]
, [b].[Column94]
, [b].[Column95]
, [b].[Column96]
, [b].[Column97]
, [b].[Column98]
, [b].[Column99]
, [b].[Column100]
, [b].[Column101]
, [b].[Column102]
, [b].[Column103]
, [b].[Column104]
, [b].[Column105]
, [b].[Column106]
, [b].[Column107]
, [b].[Column108]
, [b].[Column109]
, [b].[Column110]
, [b].[Column111]
, [b].[Column112]
, [b].[Column113]
, [b].[Column114]
, [b].[Column115]
, [b].[Column116]
, [b].[Column117]
, [b].[Column118]
, [b].[Column119]
, [b].[Column120]
FROM [#foobar] [a]
CROSS APPLY
OPENJSON([a].[jsondata], '$.member')
WITH (
[MemberFirstName] NVARCHAR(200) '$.firstname'
, [MemberLastName] NVARCHAR(200) '$.lastname'
, [MemberPhone] NVARCHAR(200) '$.phone'
, [Column1] NVARCHAR(200) '$.column1'
, [Column2] NVARCHAR(200) '$.column2'
, [Column3] NVARCHAR(200) '$.column3'
, [Column4] NVARCHAR(200) '$.column4'
, [Column5] NVARCHAR(200) '$.column5'
, [Column6] NVARCHAR(200) '$.column6'
, [Column7] NVARCHAR(200) '$.column7'
, [Column8] NVARCHAR(200) '$.column8'
, [Column9] NVARCHAR(200) '$.column9'
, [Column10] NVARCHAR(200) '$.column10'
, [Column11] NVARCHAR(200) '$.column11'
, [Column12] NVARCHAR(200) '$.column12'
, [Column13] NVARCHAR(200) '$.column13'
, [Column14] NVARCHAR(200) '$.column14'
, [Column15] NVARCHAR(200) '$.column15'
, [Column16] NVARCHAR(200) '$.column16'
, [Column17] NVARCHAR(200) '$.column17'
, [Column18] NVARCHAR(200) '$.column18'
, [Column19] NVARCHAR(200) '$.column19'
, [Column20] NVARCHAR(200) '$.column20'
, [Column21] NVARCHAR(200) '$.column21'
, [Column22] NVARCHAR(200) '$.column22'
, [Column23] NVARCHAR(200) '$.column23'
, [Column24] NVARCHAR(200) '$.column24'
, [Column25] NVARCHAR(200) '$.column25'
, [Column26] NVARCHAR(200) '$.column26'
, [Column27] NVARCHAR(200) '$.column27'
, [Column28] NVARCHAR(200) '$.column28'
, [Column29] NVARCHAR(200) '$.column29'
, [Column30] NVARCHAR(200) '$.column30'
, [Column31] NVARCHAR(200) '$.column31'
, [Column32] NVARCHAR(200) '$.column32'
, [Column33] NVARCHAR(200) '$.column33'
, [Column34] NVARCHAR(200) '$.column34'
, [Column35] NVARCHAR(200) '$.column35'
, [Column36] NVARCHAR(200) '$.column36'
, [Column37] NVARCHAR(200) '$.column37'
, [Column38] NVARCHAR(200) '$.column38'
, [Column39] NVARCHAR(200) '$.column39'
, [Column40] NVARCHAR(200) '$.column40'
, [Column41] NVARCHAR(200) '$.column41'
, [Column42] NVARCHAR(200) '$.column42'
, [Column43] NVARCHAR(200) '$.column43'
, [Column44] NVARCHAR(200) '$.column44'
, [Column45] NVARCHAR(200) '$.column45'
, [Column46] NVARCHAR(200) '$.column46'
, [Column47] NVARCHAR(200) '$.column47'
, [Column48] NVARCHAR(200) '$.column48'
, [Column49] NVARCHAR(200) '$.column49'
, [Column50] NVARCHAR(200) '$.column50'
, [Column51] NVARCHAR(200) '$.column51'
, [Column52] NVARCHAR(200) '$.column52'
, [Column53] NVARCHAR(200) '$.column53'
, [Column54] NVARCHAR(200) '$.column54'
, [Column55] NVARCHAR(200) '$.column55'
, [Column56] NVARCHAR(200) '$.column56'
, [Column57] NVARCHAR(200) '$.column57'
, [Column58] NVARCHAR(200) '$.column58'
, [Column59] NVARCHAR(200) '$.column59'
, [Column60] NVARCHAR(200) '$.column60'
, [Column61] NVARCHAR(200) '$.column61'
, [Column62] NVARCHAR(200) '$.column62'
, [Column63] NVARCHAR(200) '$.column63'
, [Column64] NVARCHAR(200) '$.column64'
, [Column65] NVARCHAR(200) '$.column65'
, [Column66] NVARCHAR(200) '$.column66'
, [Column67] NVARCHAR(200) '$.column67'
, [Column68] NVARCHAR(200) '$.column68'
, [Column69] NVARCHAR(200) '$.column69'
, [Column70] NVARCHAR(200) '$.column70'
, [Column71] NVARCHAR(200) '$.column71'
, [Column72] NVARCHAR(200) '$.column72'
, [Column73] NVARCHAR(200) '$.column73'
, [Column74] NVARCHAR(200) '$.column74'
, [Column75] NVARCHAR(200) '$.column75'
, [Column76] NVARCHAR(200) '$.column76'
, [Column77] NVARCHAR(200) '$.column77'
, [Column78] NVARCHAR(200) '$.column78'
, [Column79] NVARCHAR(200) '$.column79'
, [Column80] NVARCHAR(200) '$.column80'
, [Column81] NVARCHAR(200) '$.column81'
, [Column82] NVARCHAR(200) '$.column82'
, [Column83] NVARCHAR(200) '$.column83'
, [Column84] NVARCHAR(200) '$.column84'
, [Column85] NVARCHAR(200) '$.column85'
, [Column86] NVARCHAR(200) '$.column86'
, [Column87] NVARCHAR(200) '$.column87'
, [Column88] NVARCHAR(200) '$.column88'
, [Column89] NVARCHAR(200) '$.column89'
, [Column90] NVARCHAR(200) '$.column90'
, [Column91] NVARCHAR(200) '$.column91'
, [Column92] NVARCHAR(200) '$.column92'
, [Column93] NVARCHAR(200) '$.column93'
, [Column94] NVARCHAR(200) '$.column94'
, [Column95] NVARCHAR(200) '$.column95'
, [Column96] NVARCHAR(200) '$.column96'
, [Column97] NVARCHAR(200) '$.column97'
, [Column98] NVARCHAR(200) '$.column98'
, [Column99] NVARCHAR(200) '$.column99'
, [Column100] NVARCHAR(200) '$.column100'
, [Column101] NVARCHAR(200) '$.column101'
, [Column102] NVARCHAR(200) '$.column102'
, [Column103] NVARCHAR(200) '$.column103'
, [Column104] NVARCHAR(200) '$.column104'
, [Column105] NVARCHAR(200) '$.column105'
, [Column106] NVARCHAR(200) '$.column106'
, [Column107] NVARCHAR(200) '$.column107'
, [Column108] NVARCHAR(200) '$.column108'
, [Column109] NVARCHAR(200) '$.column109'
, [Column110] NVARCHAR(200) '$.column110'
, [Column111] NVARCHAR(200) '$.column111'
, [Column112] NVARCHAR(200) '$.column112'
, [Column113] NVARCHAR(200) '$.column113'
, [Column114] NVARCHAR(200) '$.column114'
, [Column115] NVARCHAR(200) '$.column115'
, [Column116] NVARCHAR(200) '$.column116'
, [Column117] NVARCHAR(200) '$.column117'
, [Column118] NVARCHAR(200) '$.column118'
, [Column119] NVARCHAR(200) '$.column119'
, [Column120] NVARCHAR(200) '$.column120'
) AS [b];