Ezinye izici zokuqapha kweseva ye-MS SQL. Imihlahlandlela yokusetha amafulegi okulandelela

Isibikezelo

Imvamisa, abasebenzisi, abathuthukisi nabaphathi be-MS SQL Server DBMS babhekana nezinkinga zokusebenza kwesizindalwazi noma i-DBMS iyonke, ngakho ukuqapha i-MS SQL Server kubaluleke kakhulu.
Lesi sihloko siyisengezo esihlokweni Ukusebenzisa i-Zabbix ukuze ugade i-MS SQL Server Database futhi izohlola izici ezithile zokuqapha Iseva ye-MS SQL, ikakhulukazi: indlela yokunquma ngokushesha ukuthi yiziphi izinsiza ezingekho, kanye nezincomo zokusetha amafulegi okulandela umkhondo.
Ukuze imibhalo elandelayo isebenze, udinga ukwakha i-schema ye-inf kusizindalwazi oyifunayo ngendlela elandelayo:
Ukudala i-schema ye-inf

use <имя_БД>;
go
create schema inf;

Indlela yokuthola ukuntuleka kwe-RAM

Inkomba yokuqala yokuntuleka kwe-RAM yilapho isibonelo se-MS SQL Server idla yonke i-RAM eyabelwe yona.
Ukuze wenze lokhu, dala isethulo esilandelayo se-inf.vRAM:
Idala ukubuka kwe-inf.vRAM

CREATE view [inf].[vRAM] as
select a.[TotalAvailOSRam_Mb]						--сколько свободно ОЗУ на сервере в МБ
		 , a.[RAM_Avail_Percent]					--процент свободного ОЗУ на сервере
		 , a.[Server_physical_memory_Mb]				--сколько всего ОЗУ на сервере в МБ
		 , a.[SQL_server_committed_target_Mb]			--сколько всего ОЗУ выделено под MS SQL Server в МБ
		 , a.[SQL_server_physical_memory_in_use_Mb] 		--сколько всего ОЗУ потребляет MS SQL Server в данный момент времени в МБ
		 , a.[SQL_RAM_Avail_Percent]				--поцент свободного ОЗУ для MS SQL Server относительно всего выделенного ОЗУ для MS SQL Server
		 , a.[StateMemorySQL]						--достаточно ли ОЗУ для MS SQL Server
		 , a.[SQL_RAM_Reserve_Percent]				--процент выделенной ОЗУ для MS SQL Server относительно всего ОЗУ сервера
		 --достаточно ли ОЗУ для сервера
		, (case when a.[RAM_Avail_Percent]<10 and a.[RAM_Avail_Percent]>5 and a.[TotalAvailOSRam_Mb]<8192 then 'Warning' when a.[RAM_Avail_Percent]<=5 and a.[TotalAvailOSRam_Mb]<2048 then 'Danger' else 'Normal' end) as [StateMemoryServer]
	from
	(
		select cast(a0.available_physical_memory_kb/1024.0 as int) as TotalAvailOSRam_Mb
			 , cast((a0.available_physical_memory_kb/casT(a0.total_physical_memory_kb as float))*100 as numeric(5,2)) as [RAM_Avail_Percent]
			 , a0.system_low_memory_signal_state
			 , ceiling(b.physical_memory_kb/1024.0) as [Server_physical_memory_Mb]
			 , ceiling(b.committed_target_kb/1024.0) as [SQL_server_committed_target_Mb]
			 , ceiling(a.physical_memory_in_use_kb/1024.0) as [SQL_server_physical_memory_in_use_Mb]
			 , cast(((b.committed_target_kb-a.physical_memory_in_use_kb)/casT(b.committed_target_kb as float))*100 as numeric(5,2)) as [SQL_RAM_Avail_Percent]
			 , cast((b.committed_target_kb/casT(a0.total_physical_memory_kb as float))*100 as numeric(5,2)) as [SQL_RAM_Reserve_Percent]
			 , (case when (ceiling(b.committed_target_kb/1024.0)-1024)<ceiling(a.physical_memory_in_use_kb/1024.0) then 'Warning' else 'Normal' end) as [StateMemorySQL]
		from sys.dm_os_sys_memory as a0
		cross join sys.dm_os_process_memory as a
		cross join sys.dm_os_sys_info as b
		cross join sys.dm_os_sys_memory as v
	) as a;

Khona-ke ungakwazi ukunquma ukuthi isibonelo se-MS SQL Server idla yonke inkumbulo eyabelwe yona kusetshenziswa lo mbuzo olandelayo:

select  SQL_server_physical_memory_in_use_Mb,  SQL_server_committed_target_Mb
from [inf].[vRAM];

Uma inkomba ye-SQL_server_physical_memory_in_use_Mb ihlala ingekho ngaphansi kuka-SQL_server_committed_target_Mb, kuzomele uhlole izibalo zokulinda.
Ukuze sinqume ukushoda kwe-RAM ngezibalo zokulinda, masidale ukubuka i-inf.vWaits:
Idala ukubuka kwe-inf.vWaits

CREATE view [inf].[vWaits] as
WITH [Waits] AS
    (SELECT
        [wait_type], --имя типа ожидания
        [wait_time_ms] / 1000.0 AS [WaitS],--Общее время ожидания данного типа в миллисекундах. Это время включает signal_wait_time_ms
        ([wait_time_ms] - [signal_wait_time_ms]) / 1000.0 AS [ResourceS],--Общее время ожидания данного типа в миллисекундах без signal_wait_time_ms
        [signal_wait_time_ms] / 1000.0 AS [SignalS],--Разница между временем сигнализации ожидающего потока и временем начала его выполнения
        [waiting_tasks_count] AS [WaitCount],--Число ожиданий данного типа. Этот счетчик наращивается каждый раз при начале ожидания
        100.0 * [wait_time_ms] / SUM ([wait_time_ms]) OVER() AS [Percentage],
        ROW_NUMBER() OVER(ORDER BY [wait_time_ms] DESC) AS [RowNum]
    FROM sys.dm_os_wait_stats
    WHERE [waiting_tasks_count]>0
		and [wait_type] NOT IN (
        N'BROKER_EVENTHANDLER',         N'BROKER_RECEIVE_WAITFOR',
        N'BROKER_TASK_STOP',            N'BROKER_TO_FLUSH',
        N'BROKER_TRANSMITTER',          N'CHECKPOINT_QUEUE',
        N'CHKPT',                       N'CLR_AUTO_EVENT',
        N'CLR_MANUAL_EVENT',            N'CLR_SEMAPHORE',
        N'DBMIRROR_DBM_EVENT',          N'DBMIRROR_EVENTS_QUEUE',
        N'DBMIRROR_WORKER_QUEUE',       N'DBMIRRORING_CMD',
        N'DIRTY_PAGE_POLL',             N'DISPATCHER_QUEUE_SEMAPHORE',
        N'EXECSYNC',                    N'FSAGENT',
        N'FT_IFTS_SCHEDULER_IDLE_WAIT', N'FT_IFTSHC_MUTEX',
        N'HADR_CLUSAPI_CALL',           N'HADR_FILESTREAM_IOMGR_IOCOMPLETION',
        N'HADR_LOGCAPTURE_WAIT',        N'HADR_NOTIFICATION_DEQUEUE',
        N'HADR_TIMER_TASK',             N'HADR_WORK_QUEUE',
        N'KSOURCE_WAKEUP',              N'LAZYWRITER_SLEEP',
        N'LOGMGR_QUEUE',                N'ONDEMAND_TASK_QUEUE',
        N'PWAIT_ALL_COMPONENTS_INITIALIZED',
        N'QDS_PERSIST_TASK_MAIN_LOOP_SLEEP',
        N'QDS_CLEANUP_STALE_QUERIES_TASK_MAIN_LOOP_SLEEP',
        N'REQUEST_FOR_DEADLOCK_SEARCH', N'RESOURCE_QUEUE',
        N'SERVER_IDLE_CHECK',           N'SLEEP_BPOOL_FLUSH',
        N'SLEEP_DBSTARTUP',             N'SLEEP_DCOMSTARTUP',
        N'SLEEP_MASTERDBREADY',         N'SLEEP_MASTERMDREADY',
        N'SLEEP_MASTERUPGRADED',        N'SLEEP_MSDBSTARTUP',
        N'SLEEP_SYSTEMTASK',            N'SLEEP_TASK',
        N'SLEEP_TEMPDBSTARTUP',         N'SNI_HTTP_ACCEPT',
        N'SP_SERVER_DIAGNOSTICS_SLEEP', N'SQLTRACE_BUFFER_FLUSH',
        N'SQLTRACE_INCREMENTAL_FLUSH_SLEEP',
        N'SQLTRACE_WAIT_ENTRIES',       N'WAIT_FOR_RESULTS',
        N'WAITFOR',                     N'WAITFOR_TASKSHUTDOWN',
        N'WAIT_XTP_HOST_WAIT',          N'WAIT_XTP_OFFLINE_CKPT_NEW_LOG',
        N'WAIT_XTP_CKPT_CLOSE',         N'XE_DISPATCHER_JOIN',
        N'XE_DISPATCHER_WAIT',          N'XE_TIMER_EVENT')
    )
, ress as (
	SELECT
	    [W1].[wait_type] AS [WaitType],
	    CAST ([W1].[WaitS] AS DECIMAL (16, 2)) AS [Wait_S],--Общее время ожидания данного типа в миллисекундах. Это время включает signal_wait_time_ms
	    CAST ([W1].[ResourceS] AS DECIMAL (16, 2)) AS [Resource_S],--Общее время ожидания данного типа в миллисекундах без signal_wait_time_ms
	    CAST ([W1].[SignalS] AS DECIMAL (16, 2)) AS [Signal_S],--Разница между временем сигнализации ожидающего потока и временем начала его выполнения
	    [W1].[WaitCount] AS [WaitCount],--Число ожиданий данного типа. Этот счетчик наращивается каждый раз при начале ожидания
	    CAST ([W1].[Percentage] AS DECIMAL (5, 2)) AS [Percentage],
	    CAST (([W1].[WaitS] / [W1].[WaitCount]) AS DECIMAL (16, 4)) AS [AvgWait_S],
	    CAST (([W1].[ResourceS] / [W1].[WaitCount]) AS DECIMAL (16, 4)) AS [AvgRes_S],
	    CAST (([W1].[SignalS] / [W1].[WaitCount]) AS DECIMAL (16, 4)) AS [AvgSig_S]
	FROM [Waits] AS [W1]
	INNER JOIN [Waits] AS [W2]
	    ON [W2].[RowNum] <= [W1].[RowNum]
	GROUP BY [W1].[RowNum], [W1].[wait_type], [W1].[WaitS],
	    [W1].[ResourceS], [W1].[SignalS], [W1].[WaitCount], [W1].[Percentage]
	HAVING SUM ([W2].[Percentage]) - [W1].[Percentage] < 95 -- percentage threshold
)
SELECT [WaitType]
      ,MAX([Wait_S]) as [Wait_S]
      ,MAX([Resource_S]) as [Resource_S]
      ,MAX([Signal_S]) as [Signal_S]
      ,MAX([WaitCount]) as [WaitCount]
      ,MAX([Percentage]) as [Percentage]
      ,MAX([AvgWait_S]) as [AvgWait_S]
      ,MAX([AvgRes_S]) as [AvgRes_S]
      ,MAX([AvgSig_S]) as [AvgSig_S]
  FROM ress
  group by [WaitType];

Kulokhu, unganquma ukuntuleka kwe-RAM usebenzisa lo mbuzo olandelayo:

SELECT [Percentage]
      ,[AvgWait_S]
  FROM [inf].[vWaits]
  where [WaitType] in (
    'PAGEIOLATCH_XX',
    'RESOURCE_SEMAPHORE',
    'RESOURCE_SEMAPHORE_QUERY_COMPILE'
  );

Lapha udinga ukunaka Iphesenti nezinkomba ze-AvgWait_S. Uma zibalulekile ekupheleleni kwazo, khona-ke kukhona amathuba aphezulu kakhulu okuthi isibonelo se-MS SQL Server asinayo i-RAM eyanele. Amanani abalulekile anqunywa ngawodwana kusistimu ngayinye. Nokho, ungaqala ngenkomba elandelayo: Amaphesenti>=1 futhi AvgWait_S>=0.005.
Ukukhipha izinkomba ohlelweni lokuqapha (isibonelo, i-Zabbix), ungakha le mibuzo emibili elandelayo:

1. Ayini amaphesenti ezinhlobo zokulinda ze-RAM (isamba sazo zonke izinhlobo ezinjalo zokulinda):

   select coalesce(sum([Percentage]), 0.00) as [Percentage]
   from [inf].[vWaits]
          where [WaitType] in (
              'PAGEIOLATCH_XX',
              'RESOURCE_SEMAPHORE',
               'RESOURCE_SEMAPHORE_QUERY_COMPILE'
     );
   

2. zingaki izinhlobo zokulinda ze-RAM ezithatha ngama-millisecond (inani eliphezulu lakho konke ukubambezeleka okumaphakathi kwazo zonke izinhlobo ezinjalo zokulinda):

   select coalesce(max([AvgWait_S])*1000, 0.00) as [AvgWait_MS]
   from [inf].[vWaits]
          where [WaitType] in (
              'PAGEIOLATCH_XX',
              'RESOURCE_SEMAPHORE',
               'RESOURCE_SEMAPHORE_QUERY_COMPILE'
     );
   

Ngokusekelwe ku-dynamics yamanani atholiwe alezi zinkomba ezimbili, singaphetha ngokuthi ingabe ikhona i-RAM eyanele yesibonelo se-MS SQL Server.

Indlela yokuthola umthwalo we-CPU oweqile

Ukuze ubone ukushoda kwesikhathi se-CPU, vele usebenzise ukubuka kwesistimu ye-sys.dm_os_schedulers. Lapha, uma inkomba ye-runnable_tasks_count ihlala ingaphezu kuka-1, khona-ke kunethuba elikhulu lokuthi inani lama-cores azanele ku-MS SQL Server.
Ukuze ubonise inkomba ohlelweni lokuqapha (isibonelo, i-Zabbix), ungakha isicelo esilandelayo:

select max([runnable_tasks_count]) as [runnable_tasks_count]
from sys.dm_os_schedulers
where scheduler_id<255;

Ngokusekelwe ku-dynamics yamanani atholiwe ale nkomba, singaphetha ngokuthi ingabe sikhona yini isikhathi esanele sokucubungula (inani lama-CPU cores) sesibonelo se-MS SQL Server.
Kodwa-ke, kubalulekile ukukhumbula iqiniso lokuthi imibuzo ngokwayo ingabuza imicu eminingi ngesikhathi esisodwa. Futhi kwesinye isikhathi i-optimizer ayikwazi ukulinganisa kahle ubunkimbinkimbi bombuzo ngokwawo. Khona-ke isicelo singabelwa imicu eminingi kakhulu, okuthi ngesikhathi esithile ingakwazi ukucutshungulwa kanyekanye. Futhi lokhu kubangela uhlobo lokulinda oluhambisana nokuntuleka kwesikhathi sokucubungula, nokukhula komugqa wabahleli abasebenzisa ama-CPU cores athile, okungukuthi, inkomba ye-runnable_tasks_count izokwenyuka ezimweni ezinjalo.
Kulokhu, ngaphambi kokwandisa inani lama-CPU cores, udinga ukulungisa kahle izici zokufana zesibonelo se-MS SQL Server uqobo, futhi kusukela kunguqulo 2016, ulungiselele kahle izakhiwo zokufana zolwazi oludingekayo:



Lapha kufanele unake amapharamitha alandelayo:

1. I-Max Degree of Parallelism—isetha inani eliphezulu lemicu enganikezwa esicelweni ngasinye (okuzenzakalelayo kungu-0—kukhawulwe isistimu yokusebenza ngokwayo kanye nohlelo lwe-MS SQL Server)
2. I-Cost Threshold for Parallelism - izindleko ezilinganiselwe zokuhambisana (okuzenzakalelayo ngu-5)
3. I-Max DOP—isetha inani eliphezulu lemicu engabelwe umbuzo ngamunye ezingeni lesizindalwazi (kodwa hhayi ngaphezu kwevelu yendawo ethi “Max Degree of Parallelism”) (ngokuzenzakalelayo ingu-0—inqunyelwe isistimu yokusebenza kuphela ngokwayo kanye nohlelo lwe-MS SQL Server, kanye nomkhawulo endaweni ethi “Max Degree of Parallelism” yaso sonke isibonelo se-MS SQL Server)

Akunakwenzeka ukunikeza iresiphi enhle ngokulinganayo kuwo wonke amacala, okungukuthi, udinga ukuhlaziya imibuzo enzima.
Ngokusekelwe kokuhlangenwe nakho kwami, ngincoma i-algorithm elandelayo yezenzo zezinhlelo ze-OLTP ukuze zilungiselele izakhiwo zokuhambisana:

1. qala ukhubaze ukufana ngokumisa i-Max Degree of Parallelism ibe ngu-1 ezingeni laso sonke isenzakalo
2. hlaziya imibuzo enzima kakhulu bese ukhetha inani elifanele lemicu yazo
3. setha i-Max Degree of Parallelism enanini eliphezulu elikhethiwe lemicu etholwe esinyathelweni sesi-2, futhi kusizindalwazi esithile setha inani le-Max DOP elitholwe esinyathelweni sesi-2 kusizindalwazi ngasinye.
4. hlaziya imibuzo enzima kakhulu futhi uhlonze umthelela ongemuhle wokufunda okuningiliziwe. Uma kunjalo, khulisa i-Cost Threshold for Parallelism.
   Kuzinhlelo ezifana ne-1C, i-Microsoft CRM ne-Microsoft NAV, ezikhathini eziningi ukuvimbela ukufunda okuningi kufanelekile.

Futhi, uma une-Standard edition, lapho-ke ezimeni eziningi ukuvinjelwa kochungechunge oluningi kufanelekile ngenxa yokuthi lolu hlobo lukhawulelwe enanini lama-CPU cores.
I-algorithm echazwe ngenhla ayiwafanele amasistimu e-OLAP.
Ngokusekelwe kokuhlangenwe nakho kwami, ngincoma i-algorithm elandelayo yezenzo zezinhlelo ze-OLAP ukuze zilungiselele izakhiwo zokuhambisana:

1. hlaziya imibuzo enzima kakhulu bese ukhetha inani elifanele lemicu yazo
2. setha i-Max Degree of Parallelism enanini eliphezulu elikhethiwe lemicu etholwe esinyathelweni sesi-1, futhi kusizindalwazi esithile setha inani le-Max DOP elitholwe esinyathelweni sesi-1 kusizindalwazi ngasinye.
3. hlaziya imibuzo enzima kakhulu futhi uhlonze umthelela ongemuhle wokukhawulela ukuhambisana kwemali. Uma kunjalo, yehlisa i-Cost Threshold ye-Parallelism value, noma phinda izinyathelo 1-2 zale algorithm.

Okusho ukuthi, kumasistimu e-OLTP sisuka ekuxhumeni okukodwa siye ekucushweni okuningi, futhi kumasistimu e-OLAP, ngokuphambene nalokho, sisuka ekucushweni okuningi siye kokukodwa. Ngale ndlela ungakhetha izilungiselelo ezifanele zokufana kokubili kusizindalwazi esithile kanye naso sonke isibonelo se-MS SQL Server.
Kubalulekile futhi ukuqonda ukuthi izilungiselelo zezakhiwo ezihambisanayo zidinga ukushintshwa ngokuhamba kwesikhathi, ngokusekelwe emiphumeleni yokuqapha ukusebenza kwe-MS SQL Server.

Izincomo zokusetha amafulegi okulandelela

Kusukela kokuhlangenwe nakho kwami ​​kanye nolwazi lozakwethu, ukuze ngisebenze kahle, ngincoma ukusetha amafulegi alandelayo ku-MS SQL Server run level level yezinguqulo zika-2008-2016:

1. 610 - Nciphisa ukungena kokufakiwe kumathebula anezikhombo. Ingasiza ngokufaka amathebula ngenani elikhulu lamarekhodi kanye nokwenziwe okuningi, okuvame ukulinda i-WRITELOG ende ukuze kube noshintsho ezinkombeni
2. 1117 - Uma ifayela eliseqenjini lefayela lihlangabezana nomkhawulo wokukhula okuzenzakalelayo, wonke amafayela kuqembu lamafayela ayakhuliswa
3. 1118 - Iphoqa zonke izinto ukuthi zibekwe ngezilinganiso ezihlukene (ayivumeli ububanzi obuxubile), okunciphisa isidingo sokuskena ikhasi le-SGAM, elisetshenziselwa ukulandelela ububanzi obuxubile.
4. 1224 - Ikhubaza ukukhuphuka kokukhiya okusekelwe ekubalweni kokukhiya. Nokho, ukusetshenziswa kwenkumbulo ngokweqile kunganika amandla ukukhuphuka kokukhiya
5. 2371 - Ishintsha umkhawulo wokubuyekeza izibalo ezizenzakalelayo ongashintshi ukuze ube umkhawulo wokubuyekeza izibalo ezizenzakalelayo. Kubalulekile ekubuyekezeni amapulani emibuzo kumathebula amakhulu lapho ukuchaza ngokungalungile inani lamarekhodi kuphumela ezinhlelweni zokwenziwa eziyiphutha
6. 3226 - Icindezela imilayezo yempumelelo yesipele kulogi yephutha
7. 4199 - Ihlanganisa izinguquko ku-optimizer yemibuzo ekhishwe ku-SQL Server update rollups namaphakethe wesevisi
8. 6532-6534 - Ifaka ukuthuthukiswa kokusebenza kwemibuzo enezinhlobo zedatha yendawo
9. 8048 - Iguqula izinto zememori ehlukaniswe i-NUMA zibe ezihlukaniswe nge-CPU
10. 8780 - Inika amandla isabelo sesikhathi esengeziwe sokuhlela imibuzo. Ezinye izicelo ezingenalo leli fulegi zinganqatshwa ngoba azinalo uhlelo lombuzo (iphutha elingandile kakhulu)
11. 8780 - 9389 - Inika amandla isilondolozi sememori yesikhashana esingeziwe esishintshashintshayo sama-opharetha emodi ye-batch, ivumela u-opharetha wemodi ye-batch ukuthi acele inkumbulo eyengeziwe futhi agweme ukudlulisa idatha ku-temdb uma imemori eyengeziwe itholakala.

Kuyasiza futhi ukunika amandla ifulegi lokulandela umkhondo 2016 ngaphambi kwenguqulo 2301, okuvumela ukwenziwa ngcono kokusekela izinqumo futhi ngaleyo ndlela kusize ekukhetheni izinhlelo ezingcono zemibuzo. Kodwa-ke, kusukela enguqulweni ka-2016, ivamise ukuba nomthelela omubi ezikhathini ezinde kakhulu zokwenza imibuzo.
Futhi, kumasistimu anezinkomba eziningi (isibonelo, kusizindalwazi se-1C), ngincoma ukunika amandla ifulegi lokulandelela 2330, elikhubaza ukuqoqwa kokusetshenziswa kwenkomba, ngokuvamile okuba nomthelela omuhle kusistimu.
Ungafunda kabanzi mayelana namafulegi okulandelela lapha
Kusukela kusixhumanisi esingenhla, kubalulekile futhi ukucabangela izinguqulo nokwakhiwa kwe-MS SQL Server, njengezinguqulo ezintsha, amanye amafulegi okulandelela anikwe amandla ngokuzenzakalela noma awanawo umthelela.
Unganika amandla noma ukhubaze ifulegi lokulandelela usebenzisa imiyalo ye-DBCC TRACEON kanye ne-DBCC TRACEOFF, ngokulandelanayo. Bona imininingwane eyengeziwe lapha
Ungathola isimo sokulandela umkhondo wamafulegi usebenzisa umyalo we-DBCC TRACESTATUS: ulwazi oluningi
Ukuze ukulandelela amafulegi kufakwe ekuqaliseni ngokuzenzakalela kwesevisi ye-MS SQL Server, udinga ukuya ku-SQL Server Configuration Manager futhi wengeze lawa mafulegi okulandelela ngokusebenzisa -T kuzakhiwo zesevisi:

Imiphumela

Lesi sihloko sihlolisise izici ezithile zokuqapha i-MS SQL Server, ngosizo ongakwazi ukuhlonza ngokushesha ukushoda kwe-RAM nesikhathi samahhala se-CPU, kanye nenani lezinye izinkinga ezingacacile. Amafulegi okulandelela asetshenziswa kakhulu abuyekeziwe.

Imithombo:

» I-SQL Server Wait Statistics
» Izibalo zokulinda zeseva ye-SQL noma ngicela ungitshele ukuthi kubuhlungu kuphi
» Ukubuka kwesistimu sys.dm_os_schedulers
» Ukusebenzisa i-Zabbix ukuze ugade i-MS SQL Server Database
» Indlela yokuphila ye-SQL
» Landelela amafulegi
» sql.ru

Source: www.habr.com