{"id":34173,"date":"2025-11-12T07:31:35","date_gmt":"2025-11-12T04:31:35","guid":{"rendered":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/"},"modified":"2025-11-12T07:31:35","modified_gmt":"2025-11-12T04:31:35","slug":"csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow","status":"publish","type":"post","link":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/","title":{"rendered":"CSV&#8217;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow"},"content":{"rendered":"<style>\n  \/* Mobil uyumluluk i\u00e7in \u00f6rnek CSS *\/\n  body {\n    font-family: Arial, sans-serif;\n    line-height: 1.6;\n    color: #333;\n    margin: 0 auto;\n    max-width: 1200px;\n    padding: 20px;\n  }\n  h2 {\n    color: #2c3e50;\n    border-bottom: 2px solid #3498db;\n    padding-bottom: 10px;\n    margin-top: 40px;\n  }\n  h3 {\n    color: #34495e;\n    margin-top: 30px;\n  }\n  pre {\n    background-color: #ecf0f1;\n    border: 1px solid #bdc3c7;\n    padding: 15px;\n    border-radius: 5px;\n    overflow-x: auto;\n    font-size: 0.9em;\n  }\n  code {\n    font-family: 'Consolas', 'Monaco', monospace;\n  }\n  ul, ol {\n    margin-left: 20px;\n  }\n  li {\n    margin-bottom: 8px;\n  }\n  table {\n    width: 100%;\n    border-collapse: collapse;\n    margin: 20px 0;\n  }\n  th, td {\n    border: 1px solid #ddd;\n    padding: 8px;\n    text-align: left;\n  }\n  th {\n    background-color: #f2f2f2;\n  }\n  .AdviserTip {\n    background-color: #e6f7ff;\n    border-left: 4px solid #36a2eb;\n    padding: 15px;\n    margin: 20px 0;\n    border-radius: 4px;\n  }\n  @media (max-width: 768px) {\n    body {\n      padding: 10px;\n      font-size: 14px;\n    }\n    h2 {\n      font-size: 22px;\n    }\n    h3 {\n      font-size: 18px;\n    }\n    pre {\n      white-space: pre-wrap; \/* Kod bloklar\u0131n\u0131n k\u00fc\u00e7\u00fck ekranlarda kayd\u0131rmas\u0131n\u0131 engeller *\/\n      word-break: break-all;\n    }\n  }\n  @media (max-width: 480px) {\n    body {\n      padding: 5px;\n      font-size: 12px;\n    }\n    h2 {\n      font-size: 20px;\n    }\n    h3 {\n      font-size: 16px;\n    }\n    .AdviserTip {\n      margin: 10px 0;\n      padding: 10px;\n    }\n  }\n<\/style>\n<p>CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!<\/p>\n<p>\u0130\u015fte bu teknik makalede, bu sorunun cevab\u0131n\u0131 ad\u0131m ad\u0131m ke\u015ffedece\u011fiz. Python&#8217;\u0131n esnekli\u011fi, MySQL&#8217;in g\u00fcvenilirli\u011fi ve Apache Airflow&#8217;un otomasyon g\u00fcc\u00fcn\u00fc bir araya getirerek, s\u0131f\u0131rdan u\u00e7tan uca eksiksiz bir veri hatt\u0131 in\u015fa etmeyi \u00f6\u011freneceksiniz. Bu yolculukta, veri al\u0131m\u0131ndan d\u00f6n\u00fc\u015f\u00fcme, depolamadan i\u015f ak\u0131\u015f\u0131 otomasyonuna kadar her a\u015famay\u0131 pratik \u00f6rneklerle ele alacak ve kendi veri projeleriniz i\u00e7in sa\u011flam bir temel olu\u015fturman\u0131za yard\u0131mc\u0131 olaca\u011f\u0131z. Haydi ba\u015flayal\u0131m!<\/p>\n<h2>Veri Hatt\u0131 Nedir ve Neden Hayati \u00d6nem Ta\u015f\u0131r?<\/h2>\n<p>Veri hatt\u0131, ham veriyi kayna\u011f\u0131ndan al\u0131p, belirli d\u00f6n\u00fc\u015f\u00fcmlerden ge\u00e7irerek analiz veya raporlama i\u00e7in uygun bir hedefe (genellikle bir veri ambar\u0131 veya veri g\u00f6l\u00fc) ta\u015f\u0131yan bir dizi otomatikle\u015ftirilmi\u015f ad\u0131mdan olu\u015fan bir s\u00fcre\u00e7tir. Daha teknik bir ifadeyle, bir veri hatt\u0131 genellikle &#8220;ETL&#8221; (Extract, Transform, Load &#8211; \u00c7\u0131kart, D\u00f6n\u00fc\u015ft\u00fcr, Y\u00fckle) veya &#8220;ELT&#8221; (Extract, Load, Transform &#8211; \u00c7\u0131kart, Y\u00fckle, D\u00f6n\u00fc\u015ft\u00fcr) prensiplerini takip eder. Amac\u0131, farkl\u0131 sistemlerde da\u011f\u0131n\u0131k halde bulunan veriyi standartla\u015ft\u0131r\u0131lm\u0131\u015f, temizlenmi\u015f ve eri\u015filebilir bir formatta bir araya getirmektir. Bu sayede, i\u015fletmelerin veri odakl\u0131 kararlar almas\u0131, i\u015f s\u00fcre\u00e7lerini optimize etmesi ve rekabet avantaj\u0131 elde etmesi m\u00fcmk\u00fcn olur.<\/p>\n<p>Peki, bu kadar karma\u015f\u0131k g\u00f6r\u00fcnen bir yap\u0131 neden bu kadar hayati \u00f6nem ta\u015f\u0131r? Bir e-ticaret \u015firketinin g\u00fcnl\u00fck sat\u0131\u015f, m\u00fc\u015fteri ve envanter verilerini d\u00fc\u015f\u00fcnd\u00fc\u011f\u00fcm\u00fcz zaman, bu veriler genellikle farkl\u0131 sistemlerde, farkl\u0131 formatlarda (CSV, Excel, JSON vb.) ve d\u00fczensiz bir yap\u0131da bulunur. E\u011fer bu veriler manuel olarak toplan\u0131r, temizlenir ve analiz edilmeye \u00e7al\u0131\u015f\u0131l\u0131rsa, bu s\u00fcre\u00e7 hem inan\u0131lmaz derecede zaman al\u0131c\u0131 hem de insan hatas\u0131na a\u00e7\u0131k olacakt\u0131r. Veri tutarl\u0131l\u0131\u011f\u0131 sa\u011flanamaz, raporlar gecikir ve y\u00f6neticiler zaman\u0131nda do\u011fru kararlar alamaz. \u0130\u015fte tam da bu noktada, iyi tasarlanm\u0131\u015f bir veri hatt\u0131 devreye girer. Veri hatt\u0131, bu s\u00fcreci tamamen otomatize ederek, verilerin do\u011fru zamanda, do\u011fru formatta ve do\u011fru hedefe ula\u015fmas\u0131n\u0131 sa\u011flar. Bu otomasyon, yaln\u0131zca zaman ve maliyet tasarrufu sa\u011flamakla kalmaz, ayn\u0131 zamanda veri kalitesini art\u0131r\u0131r ve i\u015fletmelerin \u00e7evik bir \u015fekilde hareket etmesine olanak tan\u0131r.<\/p>\n<p>Bu makalede in\u015fa edece\u011fimiz veri hatt\u0131nda Python, veri al\u0131m\u0131 ve d\u00f6n\u00fc\u015f\u00fcm\u00fc a\u015famalar\u0131nda sundu\u011fu zengin k\u00fct\u00fcphaneler (\u00f6zellikle Pandas) sayesinde bizim ana silah\u0131m\u0131z olacak. MySQL, temizlenmi\u015f ve yap\u0131land\u0131r\u0131lm\u0131\u015f verilerin g\u00fcvenli, eri\u015filebilir ve performans\u0131 y\u00fcksek bir \u015fekilde depolanaca\u011f\u0131 bir veri ambar\u0131 g\u00f6revi g\u00f6recektir. Son olarak, Apache Airflow, t\u00fcm bu ad\u0131mlar\u0131n belirlenen zamanlarda, belirli bir s\u0131rayla ve hatas\u0131z bir \u015fekilde otomatik olarak \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flayacak olan orkestrasyon motorumuz olacak. Airflow sayesinde, bir g\u00f6revin ba\u015far\u0131s\u0131z olmas\u0131 durumunda otomatik yeniden deneme, bildirimler ve i\u015f ak\u0131\u015f\u0131n\u0131n genel izlenebilirli\u011fi gibi kritik yeteneklere sahip olaca\u011f\u0131z. Bu \u00fc\u00e7l\u00fcn\u00fcn sinerjisi, g\u00fc\u00e7l\u00fc ve esnek bir veri altyap\u0131s\u0131 kurmam\u0131z\u0131 m\u00fcmk\u00fcn k\u0131lacakt\u0131r.<\/p>\n<h2>Python ile CSV Verilerini Okuma, D\u00f6n\u00fc\u015ft\u00fcrme ve Haz\u0131rlama<\/h2>\n<p>Veri hatt\u0131m\u0131z\u0131n ilk ad\u0131m\u0131, ham veriyi kaynaktan almakt\u0131r. Genellikle i\u015f s\u00fcre\u00e7lerinde kar\u015f\u0131la\u015ft\u0131\u011f\u0131m\u0131z en yayg\u0131n veri kaynaklar\u0131ndan biri CSV (Comma Separated Values) dosyalar\u0131d\u0131r. Bu b\u00f6l\u00fcmde, Python&#8217;\u0131n g\u00fc\u00e7l\u00fc veri i\u015fleme k\u00fct\u00fcphanesi Pandas&#8217;\u0131 kullanarak bir CSV dosyas\u0131n\u0131 nas\u0131l okuyaca\u011f\u0131m\u0131z\u0131, temel temizlik ve d\u00f6n\u00fc\u015f\u00fcm i\u015flemlerini nas\u0131l yapaca\u011f\u0131m\u0131z\u0131 ad\u0131m ad\u0131m inceleyece\u011fiz. Bu i\u015flemler, verilerimizi analiz ve depolama i\u00e7in uygun hale getirecektir. Bir senaryo d\u00fc\u015f\u00fcnelim: Bir \u015firket, g\u00fcnl\u00fck sipari\u015f verilerini i\u00e7eren &#8216;siparisler.csv&#8217; ad\u0131nda bir dosya olu\u015fturuyor. Bu dosyada <code>siparis_id<\/code>, <code>musteri_adi<\/code>, <code>urun_adi<\/code>, <code>adet<\/code>, <code>birim_fiyat<\/code> ve <code>siparis_tarihi<\/code> gibi bilgiler yer al\u0131yor. Ancak baz\u0131 s\u00fctunlarda bo\u015f de\u011ferler olabilir veya veri tipleri yanl\u0131\u015f olabilir.<\/p>\n<h3>Bir CSV Dosyas\u0131 Nas\u0131l Okunur?<\/h3>\n<p>\u00d6ncelikle, \u00f6rnek bir CSV dosyas\u0131 olu\u015ftural\u0131m. Kendi sisteminizde bir <code>siparisler.csv<\/code> dosyas\u0131 olu\u015fturup a\u015fa\u011f\u0131daki i\u00e7eri\u011fi ekleyebilirsiniz:<\/p>\n<pre><code>\nsiparis_id,musteri_adi,urun_adi,adet,birim_fiyat,siparis_tarihi\n1001,Ay\u015fe Y\u0131lmaz,Laptop,2,1500.00,2023-01-05\n1002,Mehmet Demir,Klavye,1,150.50,2023-01-05\n1003,Zeynep Kaya,Fare,3,50.25,2023-01-06\n1004,Ali Can,Monit\u00f6r,1,300.00,2023-01-06\n1005,Fatma G\u00fcne\u015f,Webcam,1,75.00,2023-01-07\n1006,Can Y\u0131lmaz,Kulakl\u0131k,NaN,120.00,2023-01-07\n1007,P\u0131nar Deniz,Router,,200.00,2023-01-08\n1008,Cem Ak,Yaz\u0131c\u0131,1,NaN,2023-01-08\n<\/pre>\n<p><\/code><\/p>\n<p>\u015eimdi Python ile bu dosyay\u0131 okuyal\u0131m:<\/p>\n<pre><code>\nimport pandas as pd\n\n# CSV dosyas\u0131n\u0131 oku\ntry:\n    df = pd.read_csv('siparisler.csv')\n    print(\"CSV dosyas\u0131 ba\u015far\u0131yla okundu. \u0130lk 5 sat\u0131r:\")\n    print(df.head())\n    print(\"\\nVeri tipleri ve bo\u015f de\u011ferler:\")\n    print(df.info())\nexcept FileNotFoundError:\n    print(\"Hata: 'siparisler.csv' dosyas\u0131 bulunamad\u0131. L\u00fctfen do\u011fru yolda oldu\u011funuzdan emin olun.\")\n<\/pre>\n<p><\/code><\/p>\n<p><code>df.head()<\/code> komutu DataFrame'in ilk birka\u00e7 sat\u0131r\u0131n\u0131 g\u00f6sterirken, <code>df.info()<\/code> DataFrame'in genel yap\u0131s\u0131, s\u00fctunlar\u0131, veri tipleri ve her s\u00fctundaki bo\u015f olmayan de\u011fer say\u0131lar\u0131 hakk\u0131nda bilgi verir. Bu, veri temizli\u011fine ba\u015flamadan \u00f6nce genel bir resim elde etmek i\u00e7in kritik bir ad\u0131md\u0131r.<\/p>\n<h3>Veri Temizleme ve D\u00f6n\u00fc\u015ft\u00fcrme Nas\u0131l Yap\u0131l\u0131r?<\/h3>\n<p>Ham veriler genellikle hatal\u0131, eksik veya yanl\u0131\u015f formatta olabilir. Bu y\u00fczden, depolama ve analiz \u00f6ncesi temizlik ve d\u00f6n\u00fc\u015f\u00fcm i\u015flemleri yapmak elzemdir. \u00d6rnek CSV dosyam\u0131zda eksik de\u011ferler (NaN) ve belki de yanl\u0131\u015f veri tipleri (\u00f6rne\u011fin, <code>siparis_tarihi<\/code> bir obje olarak okunabilir) bulunmaktad\u0131r. \u015eimdi bu sorunlar\u0131 ele alal\u0131m:<\/p>\n<pre><code>\n# 1. Eksik de\u011ferleri temizle veya doldur\n# 'adet' ve 'birim_fiyat' s\u00fctunlar\u0131ndaki bo\u015f de\u011ferleri 0 ile doldural\u0131m.\n# Veya daha sa\u011flam bir yakla\u015f\u0131m olarak, bu t\u00fcr sat\u0131rlar\u0131 tamamen silebiliriz.\n# Bu \u00f6rnekte, 'adet' ve 'birim_fiyat' s\u0131f\u0131r olamayaca\u011f\u0131ndan, o sat\u0131rlar\u0131 d\u00fc\u015f\u00fcrmeyi tercih edelim.\ndf.dropna(subset=['adet', 'birim_fiyat'], inplace=True)\n\n# 'urun_adi' s\u00fctunundaki bo\u015f de\u011ferleri 'Bilinmiyor' olarak doldural\u0131m.\ndf['urun_adi'].fillna('Bilinmiyor', inplace=True)\n\nprint(\"\\nBo\u015f de\u011ferler temizlendikten sonra DataFrame:\")\nprint(df.info())\n\n# 2. Veri tiplerini d\u00f6n\u00fc\u015ft\u00fcr\n# 'adet' ve 'birim_fiyat' s\u00fctunlar\u0131n\u0131n do\u011fru say\u0131sal tiplere sahip oldu\u011fundan emin olal\u0131m.\n# 'siparis_tarihi' s\u00fctununu datetime objesine d\u00f6n\u00fc\u015ft\u00fcr.\ndf['adet'] = df['adet'].astype(int)\ndf['birim_fiyat'] = df['birim_fiyat'].astype(float)\ndf['siparis_tarihi'] = pd.to_datetime(df['siparis_tarihi'])\n\n# 3. Yeni s\u00fctunlar olu\u015ftur\n# Her sipari\u015fin toplam tutar\u0131n\u0131 hesaplayan yeni bir s\u00fctun ekleyelim.\ndf['toplam_tutar'] = df['adet'] * df['birim_fiyat']\n\nprint(\"\\nD\u00f6n\u00fc\u015f\u00fcmlerden sonra DataFrame'in ilk 5 sat\u0131r\u0131:\")\nprint(df.head())\nprint(\"\\nD\u00f6n\u00fc\u015f\u00fcmlerden sonra veri tipleri:\")\nprint(df.info())\n<\/pre>\n<p><\/code><\/p>\n<p>Bu ad\u0131mlar sayesinde, ham CSV verisini temiz, tutarl\u0131 ve analiz edilebilir bir yap\u0131ya d\u00f6n\u00fc\u015ft\u00fcrd\u00fck. <code>dropna()<\/code> ile kritik eksik verileri i\u00e7eren sat\u0131rlar\u0131 sildik, <code>fillna()<\/code> ile daha az kritik eksik verileri doldurduk ve <code>astype()<\/code> ile <code>pd.to_datetime()<\/code> kullanarak veri tiplerini do\u011fru formata getirdik. Son olarak, mevcut s\u00fctunlardan yeni ve anlaml\u0131 bir s\u00fctun (<code>toplam_tutar<\/code>) t\u00fcrettik. Bu i\u015flemler, verilerin MySQL veritaban\u0131na sorunsuz bir \u015fekilde y\u00fcklenebilmesi i\u00e7in zemin haz\u0131rlar.<\/p>\n<div class=\"AdviserTip\">\n  <strong>Uzman \u0130pucu:<\/strong> B\u00fcy\u00fck CSV dosyalar\u0131yla \u00e7al\u0131\u015f\u0131rken belle\u011fi verimli kullanmak i\u00e7in <code>pd.read_csv()<\/code> fonksiyonundaki <code>dtype<\/code> parametresini kullanarak s\u00fctunlar\u0131n veri tiplerini \u00f6nceden belirleyebilirsiniz. Ayr\u0131ca, <code>chunksize<\/code> parametresi ile dosyay\u0131 par\u00e7a par\u00e7a okuyarak bellekte a\u015f\u0131r\u0131 y\u00fcklenmeyi \u00f6nleyebilirsiniz. Bu teknik, \u00f6zellikle GB'larca b\u00fcy\u00fckl\u00fckteki dosyalar i\u00e7in hayati \u00f6nem ta\u015f\u0131r ve performans\u0131 %40'a kadar art\u0131rabilir.\n<\/div>\n<p>Bu b\u00f6l\u00fcm, veri hatt\u0131m\u0131z\u0131n \"Extract\" ve \"Transform\" a\u015famalar\u0131n\u0131 kaps\u0131yor. Bir sonraki b\u00f6l\u00fcmde, bu temizlenmi\u015f ve d\u00f6n\u00fc\u015ft\u00fcr\u00fclm\u00fc\u015f verileri MySQL veritaban\u0131na nas\u0131l y\u00fckleyece\u011fimizi ele alaca\u011f\u0131z.<\/p>\n<h2>MySQL Veritaban\u0131 Altyap\u0131s\u0131n\u0131n Kurulumu ve \u015eema Tasar\u0131m\u0131<\/h2>\n<p>Verilerimizi Python ile ba\u015far\u0131yla temizleyip d\u00f6n\u00fc\u015ft\u00fcrd\u00fckten sonra, onlar\u0131 kal\u0131c\u0131 ve yap\u0131land\u0131r\u0131lm\u0131\u015f bir \u015fekilde saklayabilece\u011fimiz g\u00fcvenilir bir yere ihtiyac\u0131m\u0131z var: Bir veritaban\u0131. \u0130li\u015fkisel veritabanlar\u0131 aras\u0131nda en pop\u00fcler ve yayg\u0131n kullan\u0131lanlardan biri olan MySQL, bu ihtiyac\u0131 kar\u015f\u0131lamak i\u00e7in m\u00fckemmel bir se\u00e7imdir. Hem a\u00e7\u0131k kaynak olmas\u0131 hem de geni\u015f bir topluluk deste\u011fine sahip olmas\u0131, projelerimizde onu tercih etmemizin ba\u015fl\u0131ca nedenlerindendir.<\/p>\n<h3>Neden MySQL ve Nas\u0131l Kurulur?<\/h3>\n<p>MySQL, \u00f6zellikle yap\u0131land\u0131r\u0131lm\u0131\u015f veriler i\u00e7in tasarlanm\u0131\u015ft\u0131r ve ACID \u00f6zelliklerini (Atomicity, Consistency, Isolation, Durability) destekleyerek veri b\u00fct\u00fcnl\u00fc\u011f\u00fcn\u00fc garanti eder. Kolay y\u00f6netilebilir olmas\u0131, y\u00fcksek performans sunmas\u0131 ve \u00f6l\u00e7eklenebilirli\u011fi, onu veri ambar\u0131 veya operasyonel veritaban\u0131 olarak kullanmak i\u00e7in ideal k\u0131lar. Kurulum a\u00e7\u0131s\u0131ndan, MySQL'i sisteminize do\u011frudan kurabilir veya daha modern ve izole bir yakla\u015f\u0131m olan Docker konteynerlerini kullanabilirsiniz. E\u011fer Docker tercih ediyorsan\u0131z, a\u015fa\u011f\u0131daki komutla basit bir MySQL sunucusu aya\u011fa kald\u0131rabilirsiniz:<\/p>\n<pre><code>\ndocker run --name some-mysql -e MYSQL_ROOT_PASSWORD=my-secret-pw -p 3306:3306 -d mysql:latest\n<\/pre>\n<p><\/code><\/p>\n<p>Bu komut, <code>my-secret-pw<\/code> \u015fifresiyle bir MySQL sunucusu ba\u015flat\u0131r ve port 3306 \u00fczerinden eri\u015filebilir hale getirir.<\/p>\n<h3>Veri Modeli ve \u015eema Tasar\u0131m\u0131 Nas\u0131l Yap\u0131l\u0131r?<\/h3>\n<p>Veritaban\u0131 tasar\u0131m\u0131nda en kritik ad\u0131mlardan biri, verilerimizin nas\u0131l depolanaca\u011f\u0131n\u0131 belirleyen \u015fema tasar\u0131m\u0131d\u0131r. Do\u011fru bir \u015fema, veri b\u00fct\u00fcnl\u00fc\u011f\u00fcn\u00fc sa\u011flar, sorgu performans\u0131n\u0131 art\u0131r\u0131r ve gelecekteki de\u011fi\u015fikliklere uyum sa\u011flamay\u0131 kolayla\u015ft\u0131r\u0131r. Python ile i\u015fledi\u011fimiz sipari\u015f verileri i\u00e7in a\u015fa\u011f\u0131daki gibi bir tablo yap\u0131s\u0131 tasarlayabiliriz:<\/p>\n<table border=\"1\">\n<thead>\n<tr>\n<th>S\u00fctun Ad\u0131<\/th>\n<th>Veri Tipi<\/th>\n<th>A\u00e7\u0131klama<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><code>id<\/code><\/td>\n<td><code>INT PRIMARY KEY AUTO_INCREMENT<\/code><\/td>\n<td>Benzersiz sipari\u015f kayd\u0131 kimli\u011fi<\/td>\n<\/tr>\n<tr>\n<td><code>siparis_id<\/code><\/td>\n<td><code>VARCHAR(50)<\/code><\/td>\n<td>Orijinal sipari\u015f kimli\u011fi<\/td>\n<\/tr>\n<tr>\n<td><code>musteri_adi<\/code><\/td>\n<td><code>VARCHAR(100)<\/code><\/td>\n<td>M\u00fc\u015fterinin ad\u0131<\/td>\n<\/tr>\n<tr>\n<td><code>urun_adi<\/code><\/td>\n<td><code>VARCHAR(150)<\/code><\/td>\n<td>Sipari\u015f edilen \u00fcr\u00fcn\u00fcn ad\u0131<\/td>\n<\/tr>\n<tr>\n<td><code>adet<\/code><\/td>\n<td><code>INT<\/code><\/td>\n<td>Sipari\u015f edilen \u00fcr\u00fcn adedi<\/td>\n<\/tr>\n<tr>\n<td><code>birim_fiyat<\/code><\/td>\n<td><code>DECIMAL(10, 2)<\/code><\/td>\n<td>\u00dcr\u00fcn\u00fcn birim fiyat\u0131<\/td>\n<\/tr>\n<tr>\n<td><code>siparis_tarihi<\/code><\/td>\n<td><code>DATETIME<\/code><\/td>\n<td>Sipari\u015fin verildi\u011fi tarih ve saat<\/td>\n<\/tr>\n<tr>\n<td><code>toplam_tutar<\/code><\/td>\n<td><code>DECIMAL(10, 2)<\/code><\/td>\n<td>Sipari\u015fin toplam tutar\u0131<\/td>\n<\/tr>\n<tr>\n<td><code>yukleme_zamani<\/code><\/td>\n<td><code>DATETIME DEFAULT CURRENT_TIMESTAMP<\/code><\/td>\n<td>Kayd\u0131n veritaban\u0131na y\u00fcklendi\u011fi zaman<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Bu tablo yap\u0131s\u0131 i\u00e7in MySQL \u00fczerinde a\u015fa\u011f\u0131daki SQL komutlar\u0131n\u0131 \u00e7al\u0131\u015ft\u0131rarak veritaban\u0131 ve tabloyu olu\u015fturabiliriz. Bu i\u015flem genellikle bir veritaban\u0131 y\u00f6netim arac\u0131 (\u00f6rne\u011fin MySQL Workbench, DBeaver) veya komut sat\u0131r\u0131 arac\u0131l\u0131\u011f\u0131yla yap\u0131l\u0131r:<\/p>\n<pre><code>\n-- Veritaban\u0131n\u0131 olu\u015ftur\nCREATE DATABASE IF NOT EXISTS data_pipeline;\n\n-- Veritaban\u0131n\u0131 kullan\nUSE data_pipeline;\n\n-- Tabloyu olu\u015ftur\nCREATE TABLE IF NOT EXISTS siparisler (\n    id INT AUTO_INCREMENT PRIMARY KEY,\n    siparis_id VARCHAR(50) NOT NULL UNIQUE,\n    musteri_adi VARCHAR(100) NOT NULL,\n    urun_adi VARCHAR(150) NOT NULL,\n    adet INT NOT NULL,\n    birim_fiyat DECIMAL(10, 2) NOT NULL,\n    siparis_tarihi DATETIME NOT NULL,\n    toplam_tutar DECIMAL(10, 2) NOT NULL,\n    yukleme_zamani DATETIME DEFAULT CURRENT_TIMESTAMP\n);\n<\/pre>\n<p><\/code><\/p>\n<p><code>IF NOT EXISTS<\/code> komutu, veritaban\u0131n\u0131n veya tablonun zaten var olup olmad\u0131\u011f\u0131n\u0131 kontrol ederek tekrar \u00e7al\u0131\u015ft\u0131rmalarda hata vermesini engeller. <code>UNIQUE<\/code> k\u0131s\u0131tlamas\u0131, <code>siparis_id<\/code>'nin her zaman benzersiz olmas\u0131n\u0131 sa\u011flar, bu da veri tekrar\u0131n\u0131 \u00f6nler. <code>NOT NULL<\/code> k\u0131s\u0131tlamas\u0131 ise ilgili s\u00fctunlarda bo\u015f de\u011ferlere izin vermez. <code>DEFAULT CURRENT_TIMESTAMP<\/code>, <code>yukleme_zamani<\/code> s\u00fctununun varsay\u0131lan olarak kayd\u0131n eklendi\u011fi zaman\u0131 almas\u0131n\u0131 sa\u011flar, bu da veri y\u00fckleme takibi i\u00e7in olduk\u00e7a faydal\u0131d\u0131r.<\/p>\n<h3>Python ile MySQL Ba\u011flant\u0131s\u0131<\/h3>\n<p>Veritaban\u0131 altyap\u0131m\u0131z haz\u0131r oldu\u011funa g\u00f6re, Python uygulamam\u0131zdan bu veritaban\u0131na nas\u0131l ba\u011flanaca\u011f\u0131m\u0131z\u0131 \u00f6\u011frenmeliyiz. Bunun i\u00e7in <code>mysql-connector-python<\/code> k\u00fct\u00fcphanesini kullanaca\u011f\u0131z. E\u011fer y\u00fckl\u00fc de\u011filse, <code>pip install mysql-connector-python<\/code> komutu ile kurabilirsiniz:<\/p>\n<pre><code>\nimport mysql.connector\n\n# Veritaban\u0131 ba\u011flant\u0131 bilgileri\ndb_config = {\n    'host': 'localhost',\n    'user': 'root',\n    'password': 'my-secret-pw', # Docker komutunda belirledi\u011finiz \u015fifre\n    'database': 'data_pipeline'\n}\n\ntry:\n    # Veritaban\u0131na ba\u011flan\n    conn = mysql.connector.connect(**db_config)\n    if conn.is_connected():\n        print(\"MySQL veritaban\u0131na ba\u015far\u0131yla ba\u011flan\u0131ld\u0131!\")\n        cursor = conn.cursor()\n        cursor.execute(\"SELECT DATABASE();\")\n        db_name = cursor.fetchone()[0]\n        print(f\"Aktif veritaban\u0131: {db_name}\")\n\nexcept mysql.connector.Error as err:\n    print(f\"Veritaban\u0131 ba\u011flant\u0131 hatas\u0131: {err}\")\nfinally:\n    # Ba\u011flant\u0131y\u0131 kapat\n    if 'conn' in locals() and conn.is_connected():\n        cursor.close()\n        conn.close()\n        print(\"MySQL ba\u011flant\u0131s\u0131 kapat\u0131ld\u0131.\")\n<\/pre>\n<p><\/code><\/p>\n<p>Bu kod blo\u011fu, Python'dan MySQL veritaban\u0131na nas\u0131l ba\u011flant\u0131 kuraca\u011f\u0131n\u0131z\u0131 ve temel bir sorgu \u00e7al\u0131\u015ft\u0131rarak ba\u011flant\u0131y\u0131 test edece\u011finizi g\u00f6sterir. Ba\u011flant\u0131 nesnesi (<code>conn<\/code>) ve imle\u00e7 nesnesi (<code>cursor<\/code>), veritaban\u0131 ile etkile\u015fim kurmak i\u00e7in kullan\u0131lacak ana ara\u00e7lard\u0131r. Bir sonraki ad\u0131mda, bu ba\u011flant\u0131y\u0131 kullanarak temizledi\u011fimiz verileri veritaban\u0131na nas\u0131l y\u00fckleyece\u011fimizi detayland\u0131raca\u011f\u0131z.<\/p>\n<h2>Python'dan MySQL'e Veri Y\u00fckleme Stratejileri<\/h2>\n<p>Veri hatt\u0131m\u0131z\u0131n \"Load\" (Y\u00fckle) a\u015famas\u0131na geldik. Python ile haz\u0131rlad\u0131\u011f\u0131m\u0131z temiz ve d\u00f6n\u00fc\u015ft\u00fcr\u00fclm\u00fc\u015f verileri, MySQL veritaban\u0131m\u0131zdaki <code>siparisler<\/code> tablosuna etkili bir \u015fekilde aktarmam\u0131z gerekiyor. Bu b\u00f6l\u00fcmde, Python'\u0131n <code>mysql.connector<\/code> k\u00fct\u00fcphanesini kullanarak verileri nas\u0131l y\u00fckleyece\u011fimizi ve \u00f6zellikle b\u00fcy\u00fck veri setleri i\u00e7in performans optimizasyonu stratejilerini ele alaca\u011f\u0131z.<\/p>\n<h3>Veri Y\u00fckleme Mant\u0131\u011f\u0131: Toplu \u0130\u015flem (Batch Insert)<\/h3>\n<p>Tek tek <code>INSERT<\/code> sorgular\u0131 \u00e7al\u0131\u015ft\u0131rmak k\u00fc\u00e7\u00fck veri setleri i\u00e7in uygun olabilirken, binlerce veya milyonlarca sat\u0131r i\u00e7eren b\u00fcy\u00fck veri setleri i\u00e7in performans\u0131 olduk\u00e7a d\u00fc\u015f\u00fckt\u00fcr. Her bir <code>INSERT<\/code> sorgusu, veritaban\u0131 sunucusu ile ayr\u0131 bir ileti\u015fim ve i\u015flem gerektirir. Bu y\u00fck\u00fc azaltmak i\u00e7in, verileri toplu halde g\u00f6ndermek (batch insert) \u00e7ok daha verimli bir yakla\u015f\u0131md\u0131r. <code>mysql.connector<\/code> k\u00fct\u00fcphanesi, bu toplu i\u015flemleri <code>executemany()<\/code> metodu ile destekler.<\/p>\n<pre><code>\nimport pandas as pd\nimport mysql.connector\nfrom datetime import datetime\n\n# \u00d6rnek DataFrame (\u00f6nceki b\u00f6l\u00fcmlerde temizledi\u011fimiz ve d\u00f6n\u00fc\u015ft\u00fcrd\u00fc\u011f\u00fcm\u00fcz df'i temsil eder)\n# Ger\u00e7ek senaryoda bu df, 'Python ile CSV Verilerini Okuma, D\u00f6n\u00fc\u015ft\u00fcrme ve Haz\u0131rlama' b\u00f6l\u00fcm\u00fcnden gelmeli.\ndata = {\n    'siparis_id': ['1001', '1002', '1003', '1004', '1005'],\n    'musteri_adi': ['Ay\u015fe Y\u0131lmaz', 'Mehmet Demir', 'Zeynep Kaya', 'Ali Can', 'Fatma G\u00fcne\u015f'],\n    'urun_adi': ['Laptop', 'Klavye', 'Fare', 'Monit\u00f6r', 'Webcam'],\n    'adet': [2, 1, 3, 1, 1],\n    'birim_fiyat': [1500.00, 150.50, 50.25, 300.00, 75.00],\n    'siparis_tarihi': ['2023-01-05', '2023-01-05', '2023-01-06', '2023-01-06', '2023-01-07']\n}\ndf = pd.DataFrame(data)\n\n# D\u00f6n\u00fc\u015f\u00fcm ad\u0131mlar\u0131 (\u00f6nceki b\u00f6l\u00fcmdeki gibi)\ndf.dropna(subset=['adet', 'birim_fiyat'], inplace=True)\ndf['urun_adi'].fillna('Bilinmiyor', inplace=True)\ndf['adet'] = df['adet'].astype(int)\ndf['birim_fiyat'] = df['birim_fiyat'].astype(float)\ndf['siparis_tarihi'] = pd.to_datetime(df['siparis_tarihi'])\ndf['toplam_tutar'] = df['adet'] * df['birim_fiyat']\n\n# Veritaban\u0131 ba\u011flant\u0131 bilgileri\ndb_config = {\n    'host': 'localhost',\n    'user': 'root',\n    'password': 'my-secret-pw',\n    'database': 'data_pipeline'\n}\n\ndef load_data_to_mysql(dataframe, config):\n    conn = None\n    cursor = None\n    try:\n        conn = mysql.connector.connect(**config)\n        cursor = conn.cursor()\n\n        # INSERT sorgusu\n        insert_query = \"\"\"\n        INSERT INTO siparisler (siparis_id, musteri_adi, urun_adi, adet, birim_fiyat, siparis_tarihi, toplam_tutar)\n        VALUES (%s, %s, %s, %s, %s, %s, %s)\n        ON DUPLICATE KEY UPDATE\n            musteri_adi = VALUES(musteri_adi),\n            urun_adi = VALUES(urun_adi),\n            adet = VALUES(adet),\n            birim_fiyat = VALUES(birim_fiyat),\n            siparis_tarihi = VALUES(siparis_tarihi),\n            toplam_tutar = VALUES(toplam_tutar),\n            yukleme_zamani = CURRENT_TIMESTAMP;\n        \"\"\"\n        \n        # DataFrame'i tuple listesine d\u00f6n\u00fc\u015ft\u00fcrme\n        # S\u00fctun s\u0131ras\u0131n\u0131n INSERT sorgusu ile ayn\u0131 olmas\u0131na dikkat edin!\n        data_to_insert = [\n            (row['siparis_id'], row['musteri_adi'], row['urun_adi'], row['adet'], \n             row['birim_fiyat'], row['siparis_tarihi'].strftime('%Y-%m-%d %H:%M:%S'), \n             row['toplam_tutar'])\n            for index, row in dataframe.iterrows()\n        ]\n        \n        # Toplu ekleme i\u015flemi\n        cursor.executemany(insert_query, data_to_insert)\n        conn.commit() # De\u011fi\u015fiklikleri kaydet\n        print(f\"{len(data_to_insert)} sat\u0131r ba\u015far\u0131yla eklendi\/g\u00fcncellendi.\")\n        \n    except mysql.connector.Error as err:\n        print(f\"Veritaban\u0131 hatas\u0131: {err}\")\n        if conn:\n            conn.rollback() # Hata durumunda de\u011fi\u015fiklikleri geri al\n            print(\"\u0130\u015flem geri al\u0131nd\u0131.\")\n    finally:\n        if cursor:\n            cursor.close()\n        if conn and conn.is_connected():\n            conn.close()\n            print(\"MySQL ba\u011flant\u0131s\u0131 kapat\u0131ld\u0131.\")\n\n# Verileri MySQL'e y\u00fckle\nload_data_to_mysql(df, db_config)\n\n# Y\u00fcklenen verileri do\u011frulamak i\u00e7in (iste\u011fe ba\u011fl\u0131)\ntry:\n    conn = mysql.connector.connect(**db_config)\n    cursor = conn.cursor()\n    cursor.execute(\"SELECT COUNT(*) FROM siparisler;\")\n    count = cursor.fetchone()[0]\n    print(f\"\\nVeritaban\u0131ndaki toplam sipari\u015f say\u0131s\u0131: {count}\")\nexcept mysql.connector.Error as err:\n    print(f\"Do\u011frulama hatas\u0131: {err}\")\nfinally:\n    if 'conn' in locals() and conn.is_connected():\n        cursor.close()\n        conn.close()\n<\/pre>\n<p><\/code><\/p>\n<h3>Performans Optimizasyonu ve Hata Y\u00f6netimi<\/h3>\n<p>Yukar\u0131daki kodda <code>ON DUPLICATE KEY UPDATE<\/code> ifadesini kulland\u0131k. Bu, <code>siparis_id<\/code> s\u00fctunu <code>UNIQUE<\/code> oldu\u011fu i\u00e7in, ayn\u0131 <code>siparis_id<\/code>'ye sahip bir kay\u0131t gelirse yeni bir sat\u0131r eklemek yerine mevcut sat\u0131r\u0131 g\u00fcncellemesini sa\u011flar. Bu, idempotency (ayn\u0131 i\u015flemin birden fazla kez \u00e7al\u0131\u015ft\u0131r\u0131lsa bile ayn\u0131 sonucu vermesi) i\u00e7in \u00f6nemlidir ve veri hatt\u0131n\u0131z\u0131n hataya dayan\u0131kl\u0131l\u0131\u011f\u0131n\u0131 art\u0131r\u0131r.<\/p>\n<ul>\n<li><strong>Toplu Boyutland\u0131rma (Batch Sizing):<\/strong> <code>executemany()<\/code> ile g\u00f6nderilecek veri blo\u011funun boyutu \u00f6nemlidir. \u00c7ok b\u00fcy\u00fck bloklar bellek sorunlar\u0131na yol a\u00e7abilirken, \u00e7ok k\u00fc\u00e7\u00fck bloklar performans\u0131 d\u00fc\u015f\u00fcrebilir. Genellikle 1000-10000 sat\u0131rl\u0131k bloklar iyi bir ba\u015flang\u0131\u00e7 noktas\u0131d\u0131r. Pandas DataFrame'i par\u00e7alara b\u00f6lerek bu kontrol\u00fc sa\u011flayabilirsiniz.<\/li>\n<li><strong>\u0130\u015flem Y\u00f6netimi (Transaction Management):<\/strong> Kodda g\u00f6rd\u00fc\u011f\u00fcn\u00fcz gibi, <code>conn.commit()<\/code> ile de\u011fi\u015fiklikleri onayl\u0131yor ve <code>conn.rollback()<\/code> ile bir hata durumunda de\u011fi\u015fiklikleri geri al\u0131yoruz. Bu, veri b\u00fct\u00fcnl\u00fc\u011f\u00fcn\u00fc korumak i\u00e7in hayati \u00f6nem ta\u015f\u0131r. E\u011fer toplu y\u00fcklemenin ortas\u0131nda bir hata olursa, t\u00fcm i\u015flem geri al\u0131narak veritaban\u0131nda tutars\u0131z bir durum olu\u015fmas\u0131 engellenir.<\/li>\n<li><strong>\u0130ndeksleme:<\/strong> Veritaban\u0131 taraf\u0131nda, s\u0131k\u00e7a arama yap\u0131lacak veya birle\u015ftirme yap\u0131lacak s\u00fctunlara (\u00f6rne\u011fin <code>siparis_id<\/code>) uygun indeksler eklemek, <code>INSERT<\/code> i\u015flemlerini yava\u015flatabilse de, genel sorgu performans\u0131n\u0131 b\u00fcy\u00fck \u00f6l\u00e7\u00fcde art\u0131r\u0131r. <code>ON DUPLICATE KEY UPDATE<\/code> durumunda, benzersiz indeksin varl\u0131\u011f\u0131 zaten zorunludur.<\/li>\n<li><strong>Veri Tipi E\u015fle\u015fmesi:<\/strong> Python'dan g\u00f6nderilen verilerin tiplerinin MySQL tablosundaki s\u00fctun tipleriyle e\u015fle\u015fti\u011finden emin olun. Tarih ve saat gibi \u00f6zel tipler i\u00e7in <code>strftime('%Y-%m-%d %H:%M:%S')<\/code> kullanarak do\u011fru formatta g\u00f6ndermek \u00f6nemlidir.<\/li>\n<\/ul>\n<div class=\"AdviserTip\">\n  <strong>Uzman \u0130pucu:<\/strong> B\u00fcy\u00fck veri y\u00fcklemelerinde, <code>LOAD DATA INFILE<\/code> gibi MySQL'in yerel toplu y\u00fckleme ara\u00e7lar\u0131 genellikle Python tabanl\u0131 <code>executemany<\/code>'den bile daha h\u0131zl\u0131d\u0131r. E\u011fer performans kritikse ve dosya eri\u015fimi m\u00fcmk\u00fcnse, bu y\u00f6ntemi de\u011ferlendirebilirsiniz. Ancak esneklik ve veri d\u00f6n\u00fc\u015f\u00fcm\u00fc ihtiyac\u0131 varsa, Python ve <code>executemany<\/code> daha uygun bir se\u00e7enek olabilir.\n<\/div>\n<p>Bu b\u00f6l\u00fcm, veri hatt\u0131m\u0131z\u0131n en \u00f6nemli ad\u0131mlar\u0131ndan biri olan \"Load\" a\u015famas\u0131n\u0131 tamaml\u0131yor. Verilerimiz art\u0131k g\u00fcvenilir bir \u015fekilde MySQL veritaban\u0131m\u0131zda bulunuyor. Bir sonraki ad\u0131mda, t\u00fcm bu i\u015flemleri otomatikle\u015ftirerek d\u00fczenli aral\u0131klarla \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flayacak olan Apache Airflow'u devreye sokaca\u011f\u0131z.<\/p>\n<h2>Apache Airflow ile Veri Hatt\u0131n\u0131 Otomatize Etme<\/h2>\n<p>Verilerimizi okuduk, temizledik, d\u00f6n\u00fc\u015ft\u00fcrd\u00fck ve MySQL'e y\u00fckledik. Ancak bu i\u015flemlerin her g\u00fcn veya belirli aral\u0131klarla manuel olarak \u00e7al\u0131\u015ft\u0131r\u0131lmas\u0131 ne kadar s\u00fcrd\u00fcr\u00fclebilir? \u0130\u015fte bu noktada Apache Airflow devreye giriyor. Airflow, karma\u015f\u0131k veri i\u015f ak\u0131\u015flar\u0131n\u0131 programatik olarak tan\u0131mlaman\u0131za, zamanlaman\u0131za, izlemenize ve y\u00f6netmenize olanak tan\u0131yan a\u00e7\u0131k kaynakl\u0131 bir platformdur. Veri hatt\u0131m\u0131zdaki her ad\u0131m\u0131 bir g\u00f6rev (Task) olarak tan\u0131mlayacak ve bu g\u00f6revler aras\u0131ndaki ba\u011f\u0131ml\u0131l\u0131klar\u0131 belirleyerek bir i\u015f ak\u0131\u015f\u0131 (DAG - Directed Acyclic Graph) olu\u015fturaca\u011f\u0131z.<\/p>\n<h3>Airflow Nedir ve Neden Kullanmal\u0131y\u0131z?<\/h3>\n<p>Airflow, veri m\u00fchendisleri i\u00e7in bir \"i\u015f ak\u0131\u015f\u0131 orkestrasyonu\" arac\u0131d\u0131r. Geleneksel <code>cron<\/code> i\u015flerinin veya \u00f6zel scriptlerin \u00f6tesine ge\u00e7erek, a\u015fa\u011f\u0131daki avantajlar\u0131 sunar:<\/p>\n<ul>\n<li><strong>Programatik \u0130\u015f Ak\u0131\u015flar\u0131:<\/strong> \u0130\u015f ak\u0131\u015flar\u0131 (DAG'ler) Python kodunda tan\u0131mlan\u0131r, bu da s\u00fcr\u00fcm kontrol\u00fc, test edilebilirlik ve esneklik sa\u011flar.<\/li>\n<li><strong>Dinamik DAG Olu\u015fturma:<\/strong> \u015eablonlar ve kod ile dinamik olarak DAG'ler olu\u015fturulabilir.<\/li>\n<li><strong>Zengin Kullan\u0131c\u0131 Aray\u00fcz\u00fc (UI):<\/strong> T\u00fcm i\u015f ak\u0131\u015flar\u0131n\u0131, g\u00f6rev durumlar\u0131n\u0131, loglar\u0131 ve ge\u00e7mi\u015fi g\u00f6rsel olarak izlemeyi sa\u011flar.<\/li>\n<li><strong>\u00d6l\u00e7eklenebilirlik:<\/strong> Celery veya Kubernetes Executor gibi da\u011f\u0131t\u0131k mimarilerle binlerce g\u00f6revi ayn\u0131 anda \u00e7al\u0131\u015ft\u0131rabilir.<\/li>\n<li><strong>Geni\u015fletilebilirlik:<\/strong> Operat\u00f6rler, sens\u00f6rler ve kancalar arac\u0131l\u0131\u011f\u0131yla \u00e7e\u015fitli platformlar ve hizmetlerle (AWS, GCP, Azure, Spark, Hive vb.) entegre olabilir.<\/li>\n<li><strong>Hata Y\u00f6netimi ve Yeniden Denemeler:<\/strong> G\u00f6rev seviyesinde yeniden deneme politikalar\u0131, hata bildirimleri ve izleme imkanlar\u0131 sunar.<\/li>\n<\/ul>\n<p>K\u0131sacas\u0131, Airflow, veri hatt\u0131n\u0131z\u0131 tek bir noktadan y\u00f6netmenizi, potansiyel sorunlar\u0131 erken tespit etmenizi ve veri ak\u0131\u015f\u0131n\u0131z\u0131n s\u00fcrekli ve g\u00fcvenilir olmas\u0131n\u0131 sa\u011flaman\u0131z\u0131 m\u00fcmk\u00fcn k\u0131lar.<\/p>\n<h3>Temel Airflow Kavramlar\u0131<\/h3>\n<ul>\n<li><strong>DAG (Directed Acyclic Graph):<\/strong> Tan\u0131mlad\u0131\u011f\u0131n\u0131z t\u00fcm i\u015f ak\u0131\u015f\u0131n\u0131n ana yap\u0131s\u0131d\u0131r. Y\u00f6nl\u00fc (i\u015f ak\u0131\u015f\u0131n\u0131n belirli bir s\u0131ras\u0131 vard\u0131r), d\u00f6ng\u00fcsel olmayan (g\u00f6revler aras\u0131nda sonsuz bir d\u00f6ng\u00fc olu\u015fmaz) bir grafik \u015feklinde ifade edilir.<\/li>\n<li><strong>Task (G\u00f6rev):<\/strong> Bir DAG i\u00e7indeki temel birimdir ve belirli bir i\u015fi (\u00f6rne\u011fin, bir Python fonksiyonunu \u00e7al\u0131\u015ft\u0131rmak, bir SQL sorgusu y\u00fcr\u00fctmek) temsil eder.<\/li>\n<li><strong>Operator (Operat\u00f6r):<\/strong> Bir g\u00f6revi tan\u0131mlamak i\u00e7in kullan\u0131lan s\u0131n\u0131ft\u0131r. \u00d6rne\u011fin, <code>PythonOperator<\/code> Python fonksiyonlar\u0131n\u0131 \u00e7al\u0131\u015ft\u0131rmak i\u00e7in, <code>BashOperator<\/code> kabuk komutlar\u0131n\u0131 y\u00fcr\u00fctmek i\u00e7in kullan\u0131l\u0131r.<\/li>\n<li><strong>Sensor (Sens\u00f6r):<\/strong> Belirli bir ko\u015fulun (\u00f6rne\u011fin, bir dosyan\u0131n varl\u0131\u011f\u0131, bir veritaban\u0131 kayd\u0131n\u0131n olu\u015fmas\u0131) ger\u00e7ekle\u015fmesini bekleyen \u00f6zel bir operat\u00f6r t\u00fcr\u00fcd\u00fcr.<\/li>\n<li><strong>Executor (Y\u00fcr\u00fct\u00fcc\u00fc):<\/strong> G\u00f6revleri fiilen nerede \u00e7al\u0131\u015ft\u0131r\u0131laca\u011f\u0131n\u0131 belirleyen bile\u015fendir (Sequential, Local, Celery, Kubernetes gibi).<\/li>\n<\/ul>\n<h3>Basit Bir Airflow DAG Olu\u015fturma<\/h3>\n<p>\u015eimdi, \u00f6nceki ad\u0131mlarda yazd\u0131\u011f\u0131m\u0131z Python kodunu otomatize edecek bir Airflow DAG'i olu\u015ftural\u0131m. Bunun i\u00e7in, Pandas ile veri i\u015fleme ve MySQL'e y\u00fckleme mant\u0131\u011f\u0131n\u0131 i\u00e7eren bir Python fonksiyonu veya script haz\u0131rlamam\u0131z gerekecek. \u00d6rnek olarak, daha \u00f6nce yazd\u0131\u011f\u0131m\u0131z <code>load_data_to_mysql<\/code> fonksiyonunu Airflow DAG'i i\u00e7inde kullanabiliriz. Airflow, DAG dosyalar\u0131n\u0131 <code>dags<\/code> klas\u00f6r\u00fc i\u00e7inde arar.<\/p>\n<pre><code>\n# dags\/csv_to_mysql_dag.py\nfrom airflow import DAG\nfrom airflow.operators.python import PythonOperator\nfrom datetime import datetime\nimport pandas as pd\nimport mysql.connector\nimport os\n\n# Veritaban\u0131 ba\u011flant\u0131 bilgileri (Airflow Connections ile y\u00f6netmek daha g\u00fcvenlidir)\n# \u015eimdilik direkt burada tan\u0131mlayal\u0131m.\ndb_config = {\n    'host': 'host.docker.internal', # Docker i\u00e7inden host makinesine eri\u015fim i\u00e7in\n    'user': 'root',\n    'password': 'my-secret-pw',\n    'database': 'data_pipeline'\n}\n\n# Python scriptimizin \u00e7al\u0131\u015faca\u011f\u0131 fonksiyon\ndef process_and_load_csv_data(**kwargs):\n    # CSV dosyas\u0131n\u0131n yolu (Airflow container'\u0131 i\u00e7indeki path)\n    # Varsay\u0131lan olarak dags klas\u00f6r\u00fcn\u00fcn yan\u0131na bir 'data' klas\u00f6r\u00fc olu\u015fturulabilir.\n    csv_file_path = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'data', 'siparisler.csv')\n    \n    # CSV dosyas\u0131n\u0131 oku\n    try:\n        df = pd.read_csv(csv_file_path)\n        print(f\"CSV dosyas\u0131 {csv_file_path} ba\u015far\u0131yla okundu. {len(df)} sat\u0131r.\")\n    except FileNotFoundError:\n        print(f\"Hata: {csv_file_path} dosyas\u0131 bulunamad\u0131.\")\n        raise # Airflow g\u00f6revinin ba\u015far\u0131s\u0131z olmas\u0131 i\u00e7in hata f\u0131rlat\n\n    # Veri Temizleme ve D\u00f6n\u00fc\u015ft\u00fcrme\n    df.dropna(subset=['adet', 'birim_fiyat'], inplace=True)\n    df['urun_adi'].fillna('Bilinmiyor', inplace=True)\n    df['adet'] = df['adet'].astype(int)\n    df['birim_fiyat'] = df['birim_fiyat'].astype(float)\n    df['siparis_tarihi'] = pd.to_datetime(df['siparis_tarihi'])\n    df['toplam_tutar'] = df['adet'] * df['birim_fiyat']\n    print(f\"Veri temizleme ve d\u00f6n\u00fc\u015f\u00fcm tamamland\u0131. {len(df)} sat\u0131r kald\u0131.\")\n\n    # MySQL'e y\u00fckleme fonksiyonu\n    conn = None\n    cursor = None\n    try:\n        conn = mysql.connector.connect(**db_config)\n        cursor = conn.cursor()\n\n        insert_query = \"\"\"\n        INSERT INTO siparisler (siparis_id, musteri_adi, urun_adi, adet, birim_fiyat, siparis_tarihi, toplam_tutar)\n        VALUES (%s, %s, %s, %s, %s, %s, %s)\n        ON DUPLICATE KEY UPDATE\n            musteri_adi = VALUES(musteri_adi),\n            urun_adi = VALUES(urun_adi),\n            adet = VALUES(adet),\n            birim_fiyat = VALUES(birim_fiyat),\n            siparis_tarihi = VALUES(siparis_tarihi),\n            toplam_tutar = VALUES(toplam_tutar),\n            yukleme_zamani = CURRENT_TIMESTAMP;\n        \"\"\"\n        \n        data_to_insert = [\n            (row['siparis_id'], row['musteri_adi'], row['urun_adi'], row['adet'], \n             row['birim_fiyat'], row['siparis_tarihi'].strftime('%Y-%m-%d %H:%M:%S'), \n             row['toplam_tutar'])\n            for index, row in df.iterrows()\n        ]\n        \n        if data_to_insert: # Veri varsa ekle\n            cursor.executemany(insert_query, data_to_insert)\n            conn.commit()\n            print(f\"{len(data_to_insert)} sat\u0131r ba\u015far\u0131yla eklendi\/g\u00fcncellendi.\")\n        else:\n            print(\"Eklenecek yeni veri bulunamad\u0131.\")\n        \n    except mysql.connector.Error as err:\n        print(f\"Veritaban\u0131 hatas\u0131: {err}\")\n        if conn:\n            conn.rollback()\n            print(\"\u0130\u015flem geri al\u0131nd\u0131.\")\n        raise # Airflow g\u00f6revinin ba\u015far\u0131s\u0131z olmas\u0131 i\u00e7in hata f\u0131rlat\n    finally:\n        if cursor:\n            cursor.close()\n        if conn and conn.is_connected():\n            conn.close()\n            print(\"MySQL ba\u011flant\u0131s\u0131 kapat\u0131ld\u0131.\")\n\n# DAG Tan\u0131m\u0131\nwith DAG(\n    dag_id='csv_to_mysql_data_pipeline',\n    start_date=datetime(2023, 1, 1),\n    schedule_interval='@daily', # Her g\u00fcn \u00e7al\u0131\u015facak\n    catchup=False, # Ge\u00e7mi\u015fteki ka\u00e7\u0131r\u0131lan \u00e7al\u0131\u015fmalar\u0131 \u00e7al\u0131\u015ft\u0131rma\n    tags=['data_pipeline', 'csv', 'mysql', 'python'],\n    default_args={\n        'owner': 'airflow',\n        'depends_on_past': False,\n        'email_on_failure': False,\n        'email_on_retry': False,\n        'retries': 1,\n        'retry_delay': pd.Timedelta(minutes=5),\n    }\n) as dag:\n    \n    # PythonOperator kullanarak veri i\u015fleme ve y\u00fckleme fonksiyonumuzu \u00e7a\u011f\u0131r\n    process_and_load_task = PythonOperator(\n        task_id='process_and_load_csv_to_mysql',\n        python_callable=process_and_load_csv_data,\n        provide_context=True # Task'e Airflow ba\u011flam\u0131n\u0131 sa\u011flar (kwargs ile eri\u015filebilir)\n    )\n\n    # \u0130sterseniz burada ba\u015fka g\u00f6revler de tan\u0131mlayabilir ve ba\u011f\u0131ml\u0131l\u0131klar\u0131n\u0131 kurabilirsiniz.\n    # \u00d6rne\u011fin, bir sonraki ad\u0131m bir dashboard'u g\u00fcncelleme g\u00f6revi olabilir.\n    # process_and_load_task >> update_dashboard_task\n<\/pre>\n<p><\/code><\/p>\n<p>Bu DAG dosyas\u0131, <code>process_and_load_csv_data<\/code> ad\u0131nda tek bir g\u00f6rev i\u00e7erir. Bu g\u00f6rev, daha \u00f6nce Python ile yapt\u0131\u011f\u0131m\u0131z t\u00fcm CSV okuma, d\u00f6n\u00fc\u015ft\u00fcrme ve MySQL'e y\u00fckleme i\u015flemlerini kapsar. <code>schedule_interval='@daily'<\/code> sayesinde, bu DAG her g\u00fcn otomatik olarak \u00e7al\u0131\u015facakt\u0131r. <code>default_args<\/code> i\u00e7inde hata durumunda yeniden denemeler ve bildirimler gibi parametreleri tan\u0131mlayarak i\u015f ak\u0131\u015f\u0131m\u0131z\u0131n daha sa\u011flam olmas\u0131n\u0131 sa\u011flayabiliriz.<\/p>\n<div class=\"AdviserTip\">\n  <strong>Uzman \u0130pucu:<\/strong> Veritaban\u0131 kimlik bilgileri ve di\u011fer hassas bilgiler i\u00e7in Airflow Connections \u00f6zelli\u011fini kullan\u0131n. Bu sayede, bu bilgileri kodunuzda a\u00e7\u0131k\u00e7a tutmak yerine Airflow UI \u00fczerinden g\u00fcvenli bir \u015fekilde y\u00f6netebilirsiniz. <code>mysql.connector.connect()<\/code> \u00e7a\u011fr\u0131s\u0131 yerine bir Airflow hook (\u00f6rne\u011fin <code>MySqlHook<\/code>) kullanmak da daha iyi bir uygulamad\u0131r.\n<\/div>\n<p>Airflow kurulumu ve DAG'lerin nas\u0131l deploy edilece\u011fi, genellikle Docker Compose veya Kubernetes ile yap\u0131l\u0131r. Bu DAG dosyas\u0131n\u0131 Airflow'un <code>dags<\/code> klas\u00f6r\u00fcne yerle\u015ftirdi\u011finizde, Airflow UI'da g\u00f6r\u00fcnecek ve manuel olarak veya zamanlamas\u0131na g\u00f6re \u00e7al\u0131\u015ft\u0131r\u0131labilecektir. Bu sayede, veri hatt\u0131n\u0131z tamamen otomatikle\u015fmi\u015f ve izlenebilir bir hale gelmi\u015f olur.<\/p>\n<h2>Veri Hatt\u0131n\u0131n \u0130yile\u015ftirilmesi: \u0130leri D\u00fczey Konular ve En \u0130yi Uygulamalar<\/h2>\n<p>Temel bir veri hatt\u0131n\u0131 ba\u015far\u0131yla kurmu\u015f olsan\u0131z da, ger\u00e7ek d\u00fcnya senaryolar\u0131nda bu hatt\u0131n daha sa\u011flam, performansl\u0131, g\u00fcvenilir ve s\u00fcrd\u00fcr\u00fclebilir olmas\u0131 i\u00e7in baz\u0131 ileri d\u00fczey konulara dikkat etmek gerekir. Bu b\u00f6l\u00fcmde, hata y\u00f6netimi, performans optimizasyonu, veri kalitesi kontrol\u00fc ve \u00f6l\u00e7eklenebilirlik gibi kritik ba\u015fl\u0131klar\u0131 ele alaca\u011f\u0131z.<\/p>\n<h3>Hata Y\u00f6netimi ve Yeniden Denemeler<\/h3>\n<p>Veri hatlar\u0131, a\u011f kesintileri, API s\u0131n\u0131rlamalar\u0131, veri taban\u0131 kilitlenmeleri veya beklenmedik veri formatlar\u0131 gibi bir\u00e7ok nedenle ba\u015far\u0131s\u0131z olabilir. Etkili bir hata y\u00f6netimi stratejisi, bu t\u00fcr durumlarla ba\u015fa \u00e7\u0131kmak i\u00e7in hayati \u00f6neme sahiptir.<\/p>\n<ul>\n<li><strong>Try-Except Bloklar\u0131:<\/strong> Python kodunuzda, \u00f6zellikle harici sistemlerle (dosya okuma, veritaban\u0131 ba\u011flant\u0131s\u0131, API \u00e7a\u011fr\u0131lar\u0131) etkile\u015fime girdi\u011finiz yerlerde kapsaml\u0131 <code>try-except<\/code> bloklar\u0131 kullan\u0131n. Bu, k\u00fc\u00e7\u00fck hatalar\u0131n t\u00fcm i\u015f ak\u0131\u015f\u0131n\u0131 durdurmas\u0131n\u0131 engeller ve spesifik hata mesajlar\u0131 ile sorun gidermeyi kolayla\u015ft\u0131r\u0131r.<\/li>\n<li><strong>Airflow Yeniden Deneme Politikalar\u0131:<\/strong> Airflow DAG'lerinizde <code>retries<\/code> ve <code>retry_delay<\/code> parametrelerini ayarlayarak g\u00f6revlerin belirli bir s\u00fcre sonra otomatik olarak yeniden denenmesini sa\u011flayabilirsiniz. Bu, ge\u00e7ici a\u011f kesintileri gibi anl\u0131k sorunlar i\u00e7in \u00e7ok etkilidir.<\/li>\n<li><strong>Hata Bildirimleri:<\/strong> G\u00f6rev ba\u015far\u0131s\u0131z oldu\u011funda e-posta, Slack veya PagerDuty gibi kanallar arac\u0131l\u0131\u011f\u0131yla otomatik bildirimler g\u00f6nderin. Airflow'un <code>email_on_failure<\/code> gibi parametreleri veya \u00f6zel bildirim operat\u00f6rleri ile bu sa\u011flanabilir.<\/li>\n<li><strong>Hatal\u0131 Kay\u0131t Y\u00f6netimi:<\/strong> Hatal\u0131 veya i\u015flenemeyen kay\u0131tlar\u0131 ayr\u0131 bir \"dead-letter queue\" veya \"hata tablosuna\" y\u00f6nlendirin. Bu, sa\u011fl\u0131kl\u0131 verilerin i\u015flenmeye devam etmesini sa\u011flarken, hatal\u0131 verilerin daha sonra incelenmesine olanak tan\u0131r.<\/li>\n<\/ul>\n<h3>Loglama ve \u0130zleme<\/h3>\n<p>Bir veri hatt\u0131n\u0131n sa\u011fl\u0131kl\u0131 \u00e7al\u0131\u015f\u0131p \u00e7al\u0131\u015fmad\u0131\u011f\u0131n\u0131 anlamak i\u00e7in kapsaml\u0131 loglama ve etkin bir izleme sistemi \u015fartt\u0131r.<\/p>\n<ul>\n<li><strong>Detayl\u0131 Loglama:<\/strong> Python kodunuzda <code>logging<\/code> mod\u00fcl\u00fcn\u00fc kullanarak detayl\u0131 log mesajlar\u0131 olu\u015fturun. Hangi verinin ne zaman i\u015flendi\u011fi, hangi d\u00f6n\u00fc\u015f\u00fcmlerin uyguland\u0131\u011f\u0131 ve olas\u0131 hatalar loglara kaydedilmelidir.<\/li>\n<li><strong>Airflow Loglar\u0131:<\/strong> Airflow UI, her g\u00f6revin \u00e7\u0131kt\u0131 loglar\u0131na eri\u015fim sa\u011flar. Bu loglar\u0131 d\u00fczenli olarak kontrol etmek ve potansiyel sorunlar\u0131 proaktif olarak tespit etmek \u00f6nemlidir.<\/li>\n<li><strong>Performans Metrikleri:<\/strong> Veritaban\u0131 performans\u0131 (sorgu s\u00fcreleri, ba\u011flant\u0131 say\u0131s\u0131), Airflow g\u00f6rev s\u00fcreleri ve sistem kaynak kullan\u0131m\u0131 (CPU, bellek) gibi metrikleri izleyin. Prometheus, Grafana veya Datadog gibi ara\u00e7lar bu metriklerin g\u00f6rselle\u015ftirilmesinde kullan\u0131labilir.<\/li>\n<\/ul>\n<h3>Performans Optimizasyonu<\/h3>\n<p>B\u00fcy\u00fck veri hacimleriyle \u00e7al\u0131\u015f\u0131rken performans kritik bir konudur. Veri hatt\u0131n\u0131z\u0131n h\u0131zl\u0131 ve verimli \u00e7al\u0131\u015fmas\u0131n\u0131 sa\u011flamak i\u00e7in \u015funlar\u0131 g\u00f6z \u00f6n\u00fcnde bulundurun:<\/p>\n<ul>\n<li><strong>Veritaban\u0131 \u0130ndeksleri:<\/strong> MySQL tablolar\u0131n\u0131zda, <code>WHERE<\/code> ko\u015fullar\u0131nda veya <code>JOIN<\/code> i\u015flemlerinde s\u0131k\u00e7a kullan\u0131lan s\u00fctunlara do\u011fru indeksleri ekleyin. Bu, sorgu s\u00fcrelerini \u00f6nemli \u00f6l\u00e7\u00fcde azaltabilir.<\/li>\n<li><strong>Toplu \u0130\u015fleme:<\/strong> Verileri k\u00fc\u00e7\u00fck par\u00e7alara (batch) b\u00f6lerek i\u015fleyin ve veritaban\u0131na toplu olarak y\u00fckleyin (<code>executemany<\/code>). Bu, veritaban\u0131 \u00fczerindeki y\u00fck\u00fc azalt\u0131r ve i\u015flem s\u00fcresini k\u0131salt\u0131r.<\/li>\n<li><strong>SQL Sorgu Optimizasyonu:<\/strong> Yava\u015f \u00e7al\u0131\u015fan SQL sorgular\u0131n\u0131 belirleyin ve optimize edin. <code>EXPLAIN<\/code> komutunu kullanarak sorgular\u0131n nas\u0131l y\u00fcr\u00fct\u00fcld\u00fc\u011f\u00fcn\u00fc analiz edin.<\/li>\n<li><strong>Bellek Y\u00f6netimi:<\/strong> Pandas DataFrame'leri bellekte b\u00fcy\u00fck yer kaplayabilir. <code>df.info(memory_usage='deep')<\/code> ile bellek kullan\u0131m\u0131n\u0131 izleyin ve <code>astype<\/code> ile daha k\u00fc\u00e7\u00fck veri tiplerine d\u00f6n\u00fc\u015ft\u00fcrerek veya <code>chunksize<\/code> kullanarak b\u00fcy\u00fck dosyalar\u0131 par\u00e7a par\u00e7a i\u015fleyerek bellek kullan\u0131m\u0131n\u0131 optimize edin.<\/li>\n<li><strong>Paralelle\u015ftirme:<\/strong> Airflow, birden fazla g\u00f6revi paralel \u00e7al\u0131\u015ft\u0131rma yetene\u011fine sahiptir. Ba\u011f\u0131ms\u0131z g\u00f6revleri paralel olarak \u00e7al\u0131\u015ft\u0131rarak toplam i\u015f ak\u0131\u015f\u0131 s\u00fcresini k\u0131salt\u0131n.<\/li>\n<\/ul>\n<h3>Veri Kalitesi ve Do\u011frulama<\/h3>\n<p>Veri hatt\u0131n\u0131z\u0131n \u00e7\u0131kt\u0131lar\u0131n\u0131n do\u011fru ve g\u00fcvenilir olmas\u0131 i\u00e7in veri kalitesi kontrol\u00fc kritik \u00f6neme sahiptir.<\/p>\n<ul>\n<li><strong>Veri Profilleme:<\/strong> Gelen verinin yap\u0131s\u0131n\u0131, da\u011f\u0131l\u0131m\u0131n\u0131 ve potansiyel anormalliklerini anlamak i\u00e7in veri profilleme ara\u00e7lar\u0131 kullan\u0131n (\u00f6rne\u011fin, Pandas Profiling).<\/li>\n<li><strong>Beklenti Tabanl\u0131 Do\u011frulama:<\/strong> Verilerin belirli kurallara (\u00f6rne\u011fin, \"siparis_id her zaman benzersiz olmal\u0131\", \"adet > 0 olmal\u0131\") uygun olup olmad\u0131\u011f\u0131n\u0131 kontrol etmek i\u00e7in Great Expectations gibi k\u00fct\u00fcphaneleri kullan\u0131n. Bu kontrolleri Airflow g\u00f6revleri aras\u0131na entegre edebilirsiniz.<\/li>\n<li><strong>Veri Anormallik Tespiti:<\/strong> Zaman i\u00e7indeki veri trendlerinden sapan ani de\u011fi\u015fimleri tespit etmek i\u00e7in basit istatistiksel y\u00f6ntemler veya makine \u00f6\u011frenimi modelleri kullan\u0131n.<\/li>\n<\/ul>\n<h3>G\u00fcvenlik ve \u00d6l\u00e7eklenebilirlik<\/h3>\n<p>Veri hatt\u0131n\u0131z\u0131n uzun vadeli ba\u015far\u0131s\u0131 i\u00e7in g\u00fcvenlik ve \u00f6l\u00e7eklenebilirlik konular\u0131 da \u00f6nemlidir.<\/p>\n<ul>\n<li><strong>Kimlik Bilgileri Y\u00f6netimi:<\/strong> Veritaban\u0131 \u015fifreleri, API anahtarlar\u0131 gibi hassas bilgileri kodda a\u00e7\u0131k\u00e7a tutmay\u0131n. Airflow Connections, HashiCorp Vault veya AWS Secrets Manager gibi ara\u00e7lar kullanarak bunlar\u0131 g\u00fcvenli bir \u015fekilde y\u00f6netin.<\/li>\n<li><strong>Yetkilendirme:<\/strong> Veritaban\u0131nda ve di\u011fer sistemlerde en az yetki prensibini uygulay\u0131n. Veri hatt\u0131n\u0131n ihtiya\u00e7 duydu\u011fu minimum izinleri verin.<\/li>\n<li><strong>Airflow \u00d6l\u00e7eklenebilirli\u011fi:<\/strong> Daha b\u00fcy\u00fck ve karma\u015f\u0131k i\u015f y\u00fckleri i\u00e7in Airflow'un CeleryExecutor veya KubernetesExecutor gibi da\u011f\u0131t\u0131k y\u00fcr\u00fct\u00fcc\u00fclerini kullanmay\u0131 d\u00fc\u015f\u00fcn\u00fcn. Bu, Airflow'un i\u015f y\u00fck\u00fcn\u00fc birden fazla worker aras\u0131nda da\u011f\u0131tmas\u0131n\u0131 sa\u011flar.<\/li>\n<li><strong>Bulut Entegrasyonu:<\/strong> B\u00fcy\u00fck veri setleri ve y\u00fcksek i\u015flem g\u00fcc\u00fc gereksinimleri i\u00e7in AWS S3, Azure Blob Storage veya Google Cloud Storage gibi bulut tabanl\u0131 depolama ve Snowflake, BigQuery gibi bulut veri ambar\u0131 \u00e7\u00f6z\u00fcmlerini de\u011ferlendirebilirsiniz. Bu platformlar genellikle yerle\u015fik \u00f6l\u00e7eklenebilirlik ve performans optimizasyonlar\u0131 sunar.<\/li>\n<\/ul>\n<div class=\"AdviserTip\">\n  <strong>Uzman \u0130pucu:<\/strong> Idempotent (tekrarlanabilir) operasyonlar tasarlamak, veri hatt\u0131n\u0131z\u0131n hataya dayan\u0131kl\u0131l\u0131\u011f\u0131n\u0131 b\u00fcy\u00fck \u00f6l\u00e7\u00fcde art\u0131r\u0131r. Bir g\u00f6revin birden fazla kez \u00e7al\u0131\u015ft\u0131r\u0131lmas\u0131 durumunda ayn\u0131 sonucu vermesini sa\u011flamak i\u00e7in <code>ON DUPLICATE KEY UPDATE<\/code> gibi SQL yap\u0131lar\u0131 veya benzersiz kimlikler \u00fczerinden upsert (insert or update) mant\u0131\u011f\u0131 kullan\u0131n.\n<\/div>\n<p>Bu ileri d\u00fczey konular\u0131 dikkate alarak ve en iyi uygulamalar\u0131 benimseyerek, CSV'den dashboard'a uzanan veri hatt\u0131n\u0131z\u0131 sadece \u00e7al\u0131\u015f\u0131r hale getirmekle kalmayacak, ayn\u0131 zamanda onu bir \u00fcretim ortam\u0131nda g\u00fcvenle kullanabilece\u011finiz, g\u00fc\u00e7l\u00fc ve s\u00fcrd\u00fcr\u00fclebilir bir yap\u0131ya d\u00f6n\u00fc\u015ft\u00fcreceksiniz.<\/p>\n<h2>Sonu\u00e7 ve S\u0131k\u00e7a Sorulan Sorular<\/h2>\n<p>Bu kapsaml\u0131 rehberde, ham CSV dosyalar\u0131ndan ba\u015flayarak, i\u015flenmi\u015f verileri bir MySQL veritaban\u0131na y\u00fckleyip, t\u00fcm s\u00fcreci Apache Airflow ile otomatikle\u015ftirerek bir dashboarda haz\u0131r hale getiren u\u00e7tan uca bir veri hatt\u0131 in\u015fa etmeyi ad\u0131m ad\u0131m \u00f6\u011frendik. Python'\u0131n g\u00fc\u00e7l\u00fc veri i\u015fleme yetenekleri (\u00f6zellikle Pandas), MySQL'in g\u00fcvenilir veri depolama \u00e7\u00f6z\u00fcmleri ve Airflow'un esnek i\u015f ak\u0131\u015f\u0131 orkestrasyonu sayesinde, karma\u015f\u0131k veri entegrasyonu ve otomasyon g\u00f6revlerini basitle\u015ftirdik. Art\u0131k bu ara\u00e7 ve tekniklerle kendi veri projelerinize g\u00fcvenle ba\u015flayabilir, manuel s\u00fcre\u00e7lerin getirdi\u011fi y\u00fck\u00fc ortadan kald\u0131rabilir ve verilerinizden daha h\u0131zl\u0131 ve g\u00fcvenilir i\u00e7g\u00f6r\u00fcler elde edebilirsiniz.<\/p>\n<p>Bir veri hatt\u0131 in\u015fa etmek sadece teknik bir zorluk de\u011fildir; ayn\u0131 zamanda veri kalitesine, hataya dayan\u0131kl\u0131l\u0131\u011fa ve \u00f6l\u00e7eklenebilirli\u011fe odaklanan stratejik bir yat\u0131r\u0131md\u0131r. Bu makalede ele ald\u0131\u011f\u0131m\u0131z ileri d\u00fczey konular ve en iyi uygulamalar, veri hatt\u0131n\u0131z\u0131 sadece \u00e7al\u0131\u015f\u0131r hale getirmekle kalmay\u0131p, onu gelece\u011fe haz\u0131r, s\u00fcrd\u00fcr\u00fclebilir bir varl\u0131k haline getirmenize yard\u0131mc\u0131 olacakt\u0131r. Unutmay\u0131n, veri d\u00fcnyas\u0131 s\u00fcrekli de\u011fi\u015fiyor ve geli\u015fiyor. Bu teknolojilere hakim olarak, veri m\u00fchendisli\u011fi kariyerinizde \u00f6nemli bir ad\u0131m atm\u0131\u015f olacaks\u0131n\u0131z. \u015eimdi, akl\u0131n\u0131za tak\u0131labilecek baz\u0131 s\u0131k\u00e7a sorulan sorulara g\u00f6z atal\u0131m:<\/p>\n<h3>S\u0131k\u00e7a Sorulan Sorular (SSS)<\/h3>\n<ul>\n<li>\n        <strong>Veri hatt\u0131 kurmak ne kadar zaman al\u0131r?<\/strong><\/p>\n<p>Projenin karma\u015f\u0131kl\u0131\u011f\u0131na, veri hacmine ve ekibin deneyimine ba\u011fl\u0131 olarak s\u00fcre de\u011fi\u015fir. Bu rehberdeki gibi basit bir CSV'den MySQL'e hatt\u0131 kurmak birka\u00e7 saatten birka\u00e7 g\u00fcne kadar s\u00fcrebilir. Daha kapsaml\u0131, birden \u00e7ok kaynaktan beslenen ve karma\u015f\u0131k d\u00f6n\u00fc\u015f\u00fcmler i\u00e7eren kurumsal bir veri hatt\u0131 ise haftalar veya aylar s\u00fcrebilir.<\/p>\n<\/li>\n<li>\n        <strong>Airflow yerine ba\u015fka bir orkestrasyon arac\u0131 kullanabilir miyim?<\/strong><\/p>\n<p>Evet, Airflow bu alandaki pop\u00fcler ara\u00e7lardan biri olsa da, Prefect, Luigi, Dagster ve Apache Nifi gibi ba\u015fka alternatifler de bulunmaktad\u0131r. Her arac\u0131n kendine \u00f6zg\u00fc g\u00fc\u00e7l\u00fc y\u00f6nleri ve kullan\u0131m senaryolar\u0131 vard\u0131r. Se\u00e7im, projenizin spesifik ihtiya\u00e7lar\u0131na, ekibinizin al\u0131\u015fk\u0131n oldu\u011fu teknolojilere ve \u00f6l\u00e7eklenebilirlik gereksinimlerinize g\u00f6re yap\u0131lmal\u0131d\u0131r.<\/p>\n<\/li>\n<li>\n        <strong>B\u00fcy\u00fck veri setleri i\u00e7in MySQL yeterli mi?<\/strong><\/p>\n<p>MySQL, binlerce veya milyonlarca kayd\u0131 depolamak ve y\u00f6netmek i\u00e7in olduk\u00e7a yeterli ve performansl\u0131 bir ili\u015fkisel veritaban\u0131d\u0131r. Ancak terabaytlarca veri veya petabaytlarca veri ile u\u011fra\u015f\u0131yorsan\u0131z ve \u00e7ok say\u0131da karma\u015f\u0131k analitik sorgu \u00e7al\u0131\u015ft\u0131r\u0131yorsan\u0131z, Snowflake, Google BigQuery, Amazon Redshift gibi \u00f6zel veri ambar\u0131 \u00e7\u00f6z\u00fcmleri veya Apache Cassandra gibi NoSQL veritabanlar\u0131 daha uygun olabilir.<\/p>\n<\/li>\n<li>\n        <strong>Veri g\u00fcvenli\u011fini nas\u0131l sa\u011flar\u0131m?<\/strong><\/p>\n<p>Veri g\u00fcvenli\u011fi, veri hatt\u0131n\u0131n her a\u015famas\u0131nda \u00f6ncelikli olmal\u0131d\u0131r. Veritaban\u0131 ba\u011flant\u0131 \u015fifreleri ve API anahtarlar\u0131 gibi hassas bilgileri Airflow Connections gibi g\u00fcvenli y\u00f6ntemlerle veya bir gizlilik y\u00f6neticisi (\u00f6rne\u011fin HashiCorp Vault) kullanarak y\u00f6netin. Veritaban\u0131 ve sunucu seviyesinde eri\u015fim kontrol\u00fc (yetkilendirme), veri \u015fifrelemesi (aktar\u0131mda ve beklemede) ve d\u00fczenli g\u00fcvenlik denetimleri de hayati \u00f6neme sahiptir. Hassas ki\u015fisel verileri maskelemek veya anonimle\u015ftirmek de bir y\u00f6ntemdir.<\/p>\n<\/li>\n<li>\n        <strong>Bu veri hatt\u0131n\u0131 nas\u0131l izlerim ve bak\u0131m\u0131n\u0131 yapar\u0131m?<\/strong><\/p>\n<p>Airflow UI, \u00e7al\u0131\u015fan g\u00f6revleri, loglar\u0131 ve olas\u0131 hatalar\u0131 izlemek i\u00e7in temel bir ara\u00e7t\u0131r. Daha geli\u015fmi\u015f izleme i\u00e7in Prometheus ve Grafana gibi ara\u00e7larla sistem metriklerini (CPU, bellek, disk kullan\u0131m\u0131) ve veritaban\u0131 performans\u0131n\u0131 izleyebilirsiniz. Bak\u0131m a\u00e7\u0131s\u0131ndan, DAG'lerinizi d\u00fczenli olarak g\u00f6zden ge\u00e7irin, ba\u011f\u0131ml\u0131l\u0131klar\u0131 g\u00fcncelleyin, Airflow'u ve kulland\u0131\u011f\u0131n\u0131z k\u00fct\u00fcphaneleri g\u00fcncel tutun ve yedekleme stratejileri uygulay\u0131n.<\/p>\n<\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m&hellip;","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"csco_page_header_type":"","csco_page_load_nextpost":"","csco_page_subscribe_form":"","csco_page_contact_form":"","footnotes":""},"categories":[138],"tags":[],"class_list":{"0":"post-34173","1":"post","2":"type-post","3":"status-publish","4":"format-standard","6":"category-mysql","7":"cs-entry","8":"cs-video-wrap"},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v20.5 (Yoast SEO v25.3.1) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>CSV&#039;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow<\/title>\n<meta name=\"description\" content=\"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\" \/>\n<meta property=\"og:locale\" content=\"tr_TR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"CSV&#039;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow\" \/>\n<meta property=\"og:description\" content=\"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!\" \/>\n<meta property=\"og:url\" content=\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\" \/>\n<meta property=\"og:site_name\" content=\"Kodlar\u0131n Gizemli D\u00fcnyas\u0131\" \/>\n<meta property=\"article:published_time\" content=\"2025-11-12T04:31:35+00:00\" \/>\n<meta name=\"author\" content=\"Fatih Soysal\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"Yazan:\" \/>\n\t<meta name=\"twitter:data1\" content=\"Fatih Soysal\" \/>\n\t<meta name=\"twitter:label2\" content=\"Tahmini okuma s\u00fcresi\" \/>\n\t<meta name=\"twitter:data2\" content=\"32 dakika\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\"},\"author\":{\"name\":\"Fatih Soysal\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"headline\":\"CSV&#8217;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow\",\"datePublished\":\"2025-11-12T04:31:35+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\"},\"wordCount\":4755,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"articleSection\":[\"MySQL\"],\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#respond\"]}],\"copyrightYear\":\"2025\",\"copyrightHolder\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#organization\"}},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\",\"name\":\"CSV'den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow\",\"isPartOf\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\"},\"datePublished\":\"2025-11-12T04:31:35+00:00\",\"description\":\"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!\",\"breadcrumb\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#breadcrumb\"},\"inLanguage\":\"tr\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/\"]}]},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Anasayfa\",\"item\":\"https:\/\/fatihsoysal.com\/blog\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"CSV&#8217;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#website\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/\",\"name\":\"Fatihsoysal.com\",\"description\":\"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim\",\"publisher\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"tr\"},{\"@type\":[\"Person\",\"Organization\"],\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1\",\"name\":\"Fatih Soysal\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"tr\",\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"contentUrl\":\"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png\",\"width\":512,\"height\":512,\"caption\":\"Fatih Soysal\"},\"logo\":{\"@id\":\"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/\"},\"description\":\"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.\",\"url\":\"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"CSV'den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow","description":"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/","og_locale":"tr_TR","og_type":"article","og_title":"CSV'den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow","og_description":"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!","og_url":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/","og_site_name":"Kodlar\u0131n Gizemli D\u00fcnyas\u0131","article_published_time":"2025-11-12T04:31:35+00:00","author":"Fatih Soysal","twitter_card":"summary_large_image","twitter_misc":{"Yazan:":"Fatih Soysal","Tahmini okuma s\u00fcresi":"32 dakika"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#article","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/"},"author":{"name":"Fatih Soysal","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"headline":"CSV&#8217;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow","datePublished":"2025-11-12T04:31:35+00:00","mainEntityOfPage":{"@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/"},"wordCount":4755,"commentCount":0,"publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"articleSection":["MySQL"],"inLanguage":"tr","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#respond"]}],"copyrightYear":"2025","copyrightHolder":{"@id":"https:\/\/fatihsoysal.com\/blog\/#organization"}},{"@type":"WebPage","@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/","url":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/","name":"CSV'den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow","isPartOf":{"@id":"https:\/\/fatihsoysal.com\/blog\/#website"},"datePublished":"2025-11-12T04:31:35+00:00","description":"CSV dosyalar\u0131ndan bir dashboarda kadar uzanan kapsaml\u0131 bir veri hatt\u0131n\u0131 Python, MySQL ve Airflow kullanarak nas\u0131l kuraca\u011f\u0131n\u0131z\u0131 ad\u0131m ad\u0131m \u00f6\u011frenin. Otomasyon ve veri m\u00fchendisli\u011fi s\u0131rlar\u0131 burada! G\u00fcn\u00fcm\u00fcz i\u015f d\u00fcnyas\u0131nda, veriler her zamankinden daha de\u011ferli hale geldi. Ancak bu verileri anlaml\u0131 i\u00e7g\u00f6r\u00fclere d\u00f6n\u00fc\u015ft\u00fcrmek, genellikle karma\u015f\u0131k ve zaman al\u0131c\u0131 bir s\u00fcre\u00e7tir. \u015eirketler, sat\u0131\u015f raporlar\u0131ndan m\u00fc\u015fteri etkile\u015fimlerine kadar farkl\u0131 formatlarda ve \u00e7e\u015fitli kaynaklardan gelen devasa veri y\u0131\u011f\u0131nlar\u0131yla m\u00fccadele ederken, manuel veri i\u015fleme y\u00f6ntemleri art\u0131k yetersiz kalmaktad\u0131r. Peki, da\u011f\u0131n\u0131k haldeki CSV dosyalar\u0131ndan ba\u015flayarak, bu verileri temizleyip i\u015fleyerek g\u00fcvenilir bir veritaban\u0131na aktarmak ve sonras\u0131nda otomatik bir \u015fekilde g\u00f6rselle\u015ftirilebilir hale getirmek m\u00fcmk\u00fcn m\u00fc? Elbette!","breadcrumb":{"@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#breadcrumb"},"inLanguage":"tr","potentialAction":[{"@type":"ReadAction","target":["https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/"]}]},{"@type":"BreadcrumbList","@id":"https:\/\/fatihsoysal.com\/blog\/csvden-dashboara-uctan-uca-veri-hatti-kurulumu-python-mysql-airflow\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Anasayfa","item":"https:\/\/fatihsoysal.com\/blog\/"},{"@type":"ListItem","position":2,"name":"CSV&#8217;den Dashboara U\u00e7tan Uca Veri Hatt\u0131 Kurulumu: Python, MySQL, Airflow"}]},{"@type":"WebSite","@id":"https:\/\/fatihsoysal.com\/blog\/#website","url":"https:\/\/fatihsoysal.com\/blog\/","name":"Fatihsoysal.com","description":"Blog - Yaz\u0131l\u0131m D\u00fcnyas\u0131 Tecr\u00fcbelerim","publisher":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/fatihsoysal.com\/blog\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"tr"},{"@type":["Person","Organization"],"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/002a254750921dcfd568a99e48240dd1","name":"Fatih Soysal","image":{"@type":"ImageObject","inLanguage":"tr","@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/","url":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","contentUrl":"https:\/\/fatihsoysal.com\/blog\/wp-content\/uploads\/2024\/04\/cropped-replicate-prediction-3kgg1hgjn5rgp0cf0p5tr0jw7w-1.png","width":512,"height":512,"caption":"Fatih Soysal"},"logo":{"@id":"https:\/\/fatihsoysal.com\/blog\/#\/schema\/person\/image\/"},"description":"Kullan\u0131m ve kodlama m\u00fckemmeliyetini odak alan uygulamalar olu\u015fturma deneyimine sahip, profesyonel olarak 15+ y\u0131l \u00fczeri deneyime sahip bir yaz\u0131l\u0131m m\u00fchendisi.","url":"https:\/\/fatihsoysal.com\/blog\/author\/fatihsoysal\/"}]}},"yoast_meta":{"yoast_wpseo_title":"","yoast_wpseo_metadesc":"","yoast_wpseo_canonical":""},"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/34173","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/comments?post=34173"}],"version-history":[{"count":0,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/posts\/34173\/revisions"}],"wp:attachment":[{"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/media?parent=34173"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/categories?post=34173"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/fatihsoysal.com\/blog\/wp-json\/wp\/v2\/tags?post=34173"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}