All scrapers can be split into 2 categories that require different infrastructure and techniques: . Good example is , it wanders around the internet, downloads all the pages it can reach and indexes them in the Google search engine. There are quite a lot of articles and books on how to do generic scraping, so I’m not going to go into details here about it. Generic scrapers Googlebot This kind of scrapers is when you need to download some pages on the website and extract some structured valuable information. Examples of such scrapers are news scrapers, competitors prices analysis, making a local copy of publicly available data (e.g. US patents). If you need this data just once then it’s fine to just write a simple script and run it on you local machine, but if the data needs to be updated regularly then you’ll soon want to make sure it runs reliably and you have observability over current system status. Targeted scrapers. Now let’s try to design a system that can execute targeted scrapers regularly. The system only has one functional requirement: it must be able to execute arbitrary scrapers on schedule. What about non-functional requirements? It must reliably schedule and execute scrapers It should not DDoS the domains being scraped It should be highly available It should properly handle stuck scrapers (infinite loops, lost worker, etc) It should provide monitoring for itself and for the scrapers being executed Here’s an example of how such a system could look like. In result we have following components: stores information about scraper schedule and parameters that are passed to the scraper logic implementation. For example this could be a static configuration file (if you don’t need to reconfigure scrapers on the fly) or standalone database. Scraper configs storage is used to store results and status of the scraper task. Scrapers jobs storage enqueues scraper jobs based on their configs. It needs to make sure that there is no other active job before enqueuing a new one. Scheduler consumes the jobs from the queue and executes the scraper logic. It also keeps job status up-to-date, for example it should mark a job as completed or failed after the logic executing is finished. Worker Now let’s see how everything works together: loads all scrapers configs and starts scheduling the jobs. Scheduler registers a job in with a status Scheduler Scrapers jobs storage Queued enqueues registered job in the Jobs queue Scheduler dequeues a scraper job from the Jobs queue and marks it as Worker Started spawns that executes scraper logic. It also starts a thread which continuously pings that the job is still active so doesn’t enqueue another job Worker Scraper runner Scheduler executes scraper logic. Scraper runner All of the requests made by the scraper logic are intercepted by which can be used for: Requests middleware Rate limiting requests to certain domains Automatically checking if the request is blocked by robots.txt Using proxy Modifying the request (e.g. adding some headers) Logging logic may save its results somewhere or call some APIs based on data it has scraped Scraper When the scraper logic is executed, the marks the job as or . Scraper runner Succeeded Failed Scraper runner How does Scraper runner execute scrapers logic? There are multiple various options. Let’s discuss how we can solve this problem in Python since most scrapers are most commonly written using it. One of the approaches that is used in is to . Basically you package your application using egg format and then Scrapyd spawns a process that loads the provided egg file and executes it. In this case we have several advantages: Scrapyd deploy eggified Python packages Packages may include dependencies in the egg file and different scrapers may use different versions of the same dependency. This can be especially useful if your system is multi-tenant. It’s possible to dynamically deploy scrapers. In this case scraper developers and platform owners are decoupled. Another approach can be used when it’s not required for the scraper developers and the platform owners to be decoupled, e.g. when only one team develops scrapers and maintains the scraping platform. In this case what can be done is to have a scraper runner implementation which imports all the scrapers implementations and then executes them based on what it has dequeued from the queue. Here’s an example implementation: # Scraper implementation interface
class ScraperABC(ABC):
    @property
    @abstractmethod
    def name(self):
        ...

    @abstractmethod
    def execute(config: Any) -> Any:
        ...

# Scraper config that is passed to the Scraper implementation
@dataclass
class ScraperConfig:
    ...

# Sample Scraper implementation
class SampleScraper(ScraperABC):
    @property
    @abstractmethod
    def name(self):
        return "sample_scraper"

    @override
    def execute(config: ScraperConfig) -> Any:
        # <do the actual scrapping>
        text = requests.get("https://example.com").text
        return text[:100]


class Worker:
    def __init__(self, scrapers: list[ScraperABC]):
        self.queue = ...
        self.scrapers = {
          scraper.name: scraper 
          for scraper in scrapers
        }

    # Scraper runner implementation
    def execute_scraper(self, name: str, config: Any) -> Any:
        if name not in self.scrapers:
            raise Exception(f"Unknown scraper {name}")
        return self.scrapers[name].execute(config)

    # Implement consuming the queue
    def start(self):
        while True:
            task = self.queue.dequeue()
            self.execute_scraper(
                task.scraper_name, 
                task.scraper_config
            )

if __name__ == "__main__":
    Worker(
        scrapers=[
            SampleScraper()
        ]
    ).start() This approach is simpler to maintain and has less caveats although it has some limitations: All of the scrapers share dependencies, which may not be a big deal since this platform is not intended to be multi-tenant. Deploying new scrapers is not dynamic and is coupled with deploying the platform which may slow down the release cycle. Availability and Scalability of the platform The availability and scalability of the platform comes down to its components. To have a bit more clear understanding on which components to choose let’s do some back-of-the-envelope calculations. Imagine we have: 1000 scrapers that run every hour Each scraper on average runs for 15 minutes (therefore we’ll have around 250 running scrapers at any point of time) Each scraper needs around 0.1 CPU (they usually do mostly IO so we don’t expect much of CPU usage) and 512 MB of memory (so in total just for running scrapers we’ll need 25 CPU and 125 GB of memory) On average 10 scrapers’ configs are changed within a day - we have mostly read traffic and almost zero write traffic: Scraper configs storage If we use static configuration then there’re no problems at all, we can just distribute the copy across all of the components. We can as well use SQL or NoSQL databases to make the system dynamic. It needs to serve around 16 RPM as each scraper needs to reload scraper config before starting executing it. - the queue only needs to serve around 16 RPM, so we shouldn’t aim for anything heavy: Job Queue We can implement a queue via RDBMS if we have one already Another option is to use distributed queues such as , , , RabbitMQ NATS Redis Kafka - given we have 250 scrapers running at any time and each of them regularly performs health check (so we can kill stuck/timed out tasks), we should have something that can handle 250 write RPS and almost zero read traffic: Scrapers jobs storage Modern RDBMS should handle this traffic without sweating Another option is to use something like Redis, as it most likely ok to not have strong durability guarantees for the scraper jobs. - each scraper job should only be scheduled by one scheduler instance, so how can we achieve this? There are couple of options: Scheduler Have one scheduler replica - if it’s ok to have some downtime for the scheduler it’s a good option since it’s the simplest one. In case we don’t tolerate the downtime we can have primary and secondary scheduler replicas. How can we ensure that secondary replica won’t be active? It can be done via acquiring a distributed lock for the right to schedule jobs. To do so we can use either RDBMS or NoSQL DB that are already used in the system. Last approach is to shard the scheduled scrapers across multiple replicas, but it only makes sense if number of scrapers is really huge and cannot fit into the memory - each worker is a stateless service so we can scale it vertically (by adding more resources) or horizontally (by adding more replicas) Worker In result the final setup can look something like this: RDBMS (e.g. MySQL) used for: Scrapers config storage Job Queue Scrapers jobs storage Scheduler global lock Scheduler + Workers Monitoring There are two parts that needs to be monitored: the platform itself and executed scrapers. Let’s start with the platform monitoring. We want to have an observability over following things: Resource utilisation: CPU, Memory, Disk usage of all the components Queue metrics: Number of pending jobs Processing lag (time since first queue item was enqueued) Average time spent in the queue Enqueue/Dequeue latency Scheduler metrics: Number of active scheduler replicas Difference between expected scheduled time and actual scheduled time (scheduler delay) Worker metrics Number of active workers replicas Difference between expected scheduled time and when the task processing started (end-to-end platform delay) Success rate of the executed tasks Number of timed out tasks All these metrics should tell us how well platform performs overall. As a platform we could also expose some metric for the scrapers developers so they can keep them up-to-date easily. Here are some metrics that come to mind: Task success rate over different periods of time - to tell if the scraper implementation is flaky or completely broken. Number of consecutive failures - this tells us that the scraper is most likely broken at all and should be fixed as soon as possible. Time since last succeeded task - this ensures we execute scrapers within SLAs (e.g. we may have a requirement to provide the data within 1 day of its releasing) Execution time - this tells us if we are able to process the data as soon as it arrives and don’t fall behind (e.g. if getting some information about 1 day period takes 2 days we’ll never be able to scrape the data up to a current point of time, hence the scraper needs to be optimised) Resource consumption (if we spawn a process for each scraper that should be fairly easy to measure) Existing solutions There are actually not that many existing solutions that provide full fledged scraping experience. The most popular is - an open source and collaborative framework for extracting the data from websites. In a fast, simple, yet extensible way. There are multiple ways to host it: Scrapy - if you’d like to host it using your own resources Scrapyd or - cloud solutions to host your spiders Scrapy Cloud ScrapeOps is another popular solution to create and run scrapers that use headless browsers such as or . Although there’s no built-in scheduling and monitoring capabilites so you’ll either need to implement them yourself or use to host the scrapers. Crawlee Playwright Selenium Apify Another notable option is to use no-code solutions like or which allows you to train scrapers by clicking on what you want to scrape. parsehub octoparse

The code in this story is for educational purposes. The readers are solely responsible for whatever they build with it.

Read My Stories

Too Long; Didn't Read

Make resilience your competitive advantage

Designing a Scraping Platform: Generic Scrapers vs. Targeted Scrapers

About Author

Comments

TOPICS

THIS ARTICLE WAS FEATURED IN

Related Stories

Untitled Story

104 Stories To Learn About Go

105 Stories To Learn About Functional Programming

100+ Free Pluralsight Courses to learn Python, Java, and Spring Boot

10 Websites to Learn JavaScript for Beginners

104 Stories To Learn About Programming Top Story

104 Stories To Learn About Go

105 Stories To Learn About Functional Programming

100+ Free Pluralsight Courses to learn Python, Java, and Spring Boot

10 Websites to Learn JavaScript for Beginners

104 Stories To Learn About Programming Top Story

Light-Mode

Classic

Newspaper

Dark-Mode

Neon Noir

Minty

HN StartUps