Awọn Itọsọna · Browser Automation

Lilo awọn proxy pẹlu Puppeteer, Playwright, ati Selenium

Awọn automated browser nilo awọn proxy lati ya awọn IP sọtọ kọja awọn session tabi lati lọ kiri lori awọn oju-iwe bi olumulo ni orilẹ-ede miiran. Iṣoro ti o wọpọ ni pe Chrome n foju fo awọn username ati password ninu argument --proxy-server, nitorinaa ile-ikawe kọọkan n pese ọna tirẹ lati ṣakoso ijerisi proxy.

Òǹkọ̀wé: Bùi Toàn · Tí a sọ di ọ̀tun: 19/11/2026

Playwright (Node.js, Python)

  • Nigbati o ba n ṣe ifilọlẹ browser kan: chromium.launch({ proxy: { server: "http://IP:PORT", username: "user", password: "pass" } }).
  • Proxy kan fun context kọọkan: browser.newContext({ proxy: { server, username, password } }). Browser instance kan le ṣiṣẹ awọn context pupọ, ọkọọkan pẹlu IP tirẹ, ti o nlo overhead ti o kere pupọ ju ṣiṣi awọn browser pupọ lọ.
  • Python nlo ilana kanna pẹlu dictionary: p.chromium.launch(proxy={"server": "http://IP:PORT", "username": "user", "password": "pass"}).
  • Firefox ati WebKit ninu Playwright gba iṣeto proxy kanna gẹlẹ.

Puppeteer

  1. Ṣe ifilọlẹ browser pẹlu command-line flag: puppeteer.launch({ args: ["--proxy-server=http://IP:PORT"] }).
  2. Fun oju-iwe kọọkan, pe await page.authenticate({ username: "user", password: "pass" }) ṣaaju page.goto.
  3. Ti o ba nilo awọn proxy pupọ laarin browser kan, lo browser.createBrowserContext({ proxyServer: "http://IP:PORT" }) ninu awọn ẹya igbalode ti Puppeteer, pẹlu page.authenticate lori oju-iwe kọọkan.
  4. Daju asopọ nipa lilọ si https://api.ipify.org ati kika response body naa.

Ṣe Selenium le lo awọn proxy pẹlu orukọ olumulo ati ọrọ igbaniwọle?

Bẹẹni, ṣugbọn kii ṣe taara. ChromeDriver ko ni paramita fun orukọ olumulo ati ọrọ igbaniwọle proxy, ati pe apoti ibaraẹnisọrọ ijerisi proxy ti Chrome ko le ṣe akoso pẹlu Selenium deede. Awọn ọna abayọ ti o wọpọ pẹlu: lilo ile-ikawe selenium-wire (Python) eyiti o gba awọn proxy ni ọna kika http://user:pass@IP:PORT; ṣiṣẹda afikun Chrome kekere kan lati mu awọn iṣẹlẹ ijẹrisi mu ati ikojọpọ rẹ ni ibẹrẹ; tabi lilo Firefox pẹlu profaili nibiti a ti fi awọn iwe-ẹri proxy pamọ si. Fun awọn iṣẹ akanṣe tuntun, Playwright mu ki eyi rọrun pupọ.

Kini o yẹ ki o ranti nigbati o ba n lo Playwright ati Puppeteer nipasẹ awọn proxy fun igba pipẹ?

Ṣeto awọn akoko ipari (timeouts) ki o tun gbiyanju lori awọn aṣiṣe nẹtiwọọki, pa awọn oju-iwe ati awọn agbegbe (contexts) nigbati o ba pari, ki o dinku awọn igba lilo akoko kan naa fun IP kan lati yago fun awọn aṣiṣe 429.

  • Ṣeto awọn navigation timeout ti o tọ (iṣẹju-aaya 30–60) ki o si fi retry logic sii fun awọn ikuna nẹtiwọọki igba diẹ.
  • Pa awọn oju-iwe ati awọn context lẹsẹkẹsẹ lẹhin ipari lati yago fun sisọnu awọn asopọ ti ko ṣiṣẹ.
  • Din awọn session nigbakanna ti n ṣiṣẹ nipasẹ IP kan ku; awọn iṣẹ ibi-afẹde le da 429 pada ti iwọn ibeere lati adirẹsi kan ba ga ju.
  • Mu akoko agbegbe ati locale ti context ba orilẹ-ede proxy mu (timezoneId, locale ninu Playwright).

Ṣe o nilo proxy? Ra awọn dedicated proxy pẹlu ifilọlẹ adaṣe

Awọn dedicated IPv4 proxy ni awọn orilẹ-ede 9 lati 20.000₫/IP/oṣu nigbati o ba n ra lati 2.000 IP (39.000₫ fun 1 IP) ati awọn static ISP proxy ni Vietnam, US. HTTP(S) ati SOCKS5 pẹlu username ati password, bandwidth ailopin, ti a pese laifọwọyi lẹhin sisanwo. Ṣe akiyesi pe awọn proxy Vietnam nlo protocol HTTP(S) lọwọlọwọ ko si tii ṣe atilẹyin SOCKS5.

Ra awọn proxy ni bayi

Àwọn Ìbéèrè Tó Wọ́pọ̀

Kini idi ti --proxy-server=http://user:pass@IP:PORT ko fi ṣiṣẹ?+

Chrome n foju fo username ati password ninu argument yii. Lo page.authenticate (Puppeteer) tabi awọn paramita proxy ti a ṣe sinu Playwright dipo.

Ṣe a le lo awọn proxy ONET pẹlu awọn headless browser?+

Bẹẹni. Awọn proxy n ṣiṣẹ ni ọna kanna boya wọn nṣiṣẹ ni headless tabi ipo GUI.