Wednesday 29 March 2017

Full day workshop(s) on StormCrawler (+Elasticsearch and Kibana)

I will be running a full-day workshop on crawling with StormCrawler on the 24th April in Berlin. See full details on

Please find the program below:

In this workshop, we will explore StormCrawler a collection of resources for building low-latency, large scale web crawlers on Apache Storm. After a short introduction to Apache Storm and an overview of what Storm-Crawler provides, we'll put it to use straight away for a simple crawl before moving on to the deployed mode of Storm

In the second part of the session, we will then introduce metrics and index documents with Elasticsearch and Kibana and dive into data extraction. Finally, we'll cover recursive crawls and scalability. This course will be hands-on: attendees will run the code on their own machines.  

This course will suit Java developers with an interest in big data, stream processing, web crawling and search. It will provide a practical introduction to both Apache Storm and Elasticsearch as well of course as StormCrawler and should not require advanced programming skills. 

Duration : 2x3 hours 

PS: Do you follow DigitalPebble or StormCrawler on Twitter? Announcements and updates are made there (as well as all sorts of interesting news of course!) 

Need billions of web pages? Don't bother crawling...

How big did you say?

I am often contacted by prospective clients to help them crawl the web on a very large scale or find questions such as this one on StackOverflow. What people want to achieve with web data varies greatly from one case to the next: some need to extract specific data from as many pages as possible, some want to build search engines, while others wish to test the accuracy of a machine learning model on real data.  

Luckily, there are resources available for large scale web crawling, both on the platform side (e.g. Amazon Web Services) and the software side (StormCrawler, Apache Nutch), however large scale crawling (think billions of pages and hundreds of servers) is costly, complex and time-consuming.  At DigitalPebble, we help our clients with such tasks but what I often tend to recommend as an initial step is to have a look at CommonCrawl.

CommonCrawl to the rescue

CommonCrawl is a non-profit organisation which provides web crawl data for free. Their datasets are used by various organisations, both in academia and industry, as can be seen on the examples page. The applications range from machine learning to natural language processing or computational linguistics. For instance, at DigitalPebble, we have used the CommonCrawl dataset for some of our clients for information extraction (phone numbers and contact details publicly available), machine learning (to check the accuracy of a classifier on real, big, messy data) as well as lexicometry (get frequencies of anchor tags). I should also mention that CommonCrawl themselves are clients of ours: we developed Apache Nutch resources for them and also ran their February 2016 web crawl. We also contributed to the set up of their news crawl (see below).

CommonCrawl provides two types of datasets, both hosted on Amazon S3 as part of the Amazon Public Datasets program.

Web crawl

The main dataset is released on a monthly basis and consists of billions of web pages stored in WARC format on AWS S3. The latest release had 3.08 billion web pages and about 250 TiB of uncompressed content: that’s a lot of data to play with, and it comes for free!

These pages are mainly HTML documents, but there are also a few PDF and images. Until recently, the coverage was very US-centric and the datasets contained mostly the same URLs from one release to the next, but this is no longer the case as European domain names and the top 1 million Alexa domains are crawled (see details on Interestingly, CommonCrawl use Apache Nutch to generate their datasets, albeit with a few home-made modifications.

Basically, each release is split into 100 segments. Each segment has three types of files WARC, WAT and WET. As explained on the Get Started page:

  • WARC files store the raw crawl data
  • WAT files store computed metadata for the data stored in the WARC
  • WET files store extracted plaintext from the data stored in the WARC

Note that WAT and WET are in the WARC format too! In fact, the WARC format is nothing more than an envelope with metadata and content. In the case of the WARC files, that content is the HTTP requests and responses, whereas for the WET files, it is simply the plain text extracted from the WARCs. The WAT files contain a JSON representation of metadata extracted from the WARCs e.g. title, links etc…

So, not only have CommonCrawl given you loads of web data for free, they’ve also made your life easier by preprocessing the data for you. For many tasks, the content of the WAT or WET files will be sufficient and you won’t have to process the WARC files.

This should not only help you simplify your code but also make the whole processing faster. We recently ran an experiment on CommonCrawl where we needed to extract anchor text from HTML pages. We initially wrote some MapReduce code to extract the binary content of the pages from their WARC representation, processed the HTML with JSoup and reduced on the anchor text. Processing a single WARC segment took roughly 100 minutes on a 10-node EMR cluster. We then simplified the extraction logic, took the WAT files as input and the processing time dropped to 17 minutes on the same cluster. This gain was partly due to not having to parse the web pages, but also to the fact that WAT files are a lot smaller than their WARC counterparts.

News dataset

Unlike the main web crawl, the news dataset is released continuously. As its name suggests, it consists exclusively of news pages and articles as described on There are between 3 and 5 WARC files (1GB each) generated daily, corresponding to 300 to 400 thousand pages. In total, over 25 million news pages have been crawled to date. The dataset contains WARC files only so you will have to write some code to extract the text and metadata yourself.

The news dataset is generated using our very own StormCrawler and the code of the news crawl is publicly available on CommonCrawl’s GitHub account.


The Get Started page on the CommonCrawl website contains useful pointers to libraries and code in various programming languages to process the datasets. There is also a list of tutorials and presentations.

It is also worth noting that CommonCrawl provides an index per release, allowing you to search for URLs (including wildcards) and retrieve the segment and offset therein where the content of the URL is stored e.g.

{ "urlkey": "org,apache)/", "timestamp": "20170220105827", "status": "200", "url": "", "filename": "crawl-data/CC-MAIN-2017-09/segments/1487501170521.30/warc/CC-MAIN-20170219104610-00206-ip-10-171-10-108.ec2.internal.warc.gz", "length": "13315", "mime": "text/html", "offset": "14131184", "digest": "KJREISJSKKGH6UX5FXGW46KROTC6MBEM" }

This is useful but only if you are interested in a limited number of URLs which you know in advance. In many cases, what you know in advance is what you want to extract, not where it will be extracted from. For situations such as these, you will need distributed batch-processing using MapReduce in Apache Hadoop or Apache Spark.

As hinted above, I tend to use AWS EMR (ElasticMapReduce). Running the code in AWS makes sense as the data sets are stored on S3 so access is fast and there is no transfer cost, also the EC2 instances will have the credentials pre-set so there is no additional configuration needed to access the data. There is an additional cost in using EMR but this saves me from having to configure Hadoop. In addition, I usually store the output of the reduce steps on a S3 bucket so that nothing is kept on HDFS and I can use spot instances to keep the cost down. If they get terminated, nothing is lost. Of course, other platforms (Azure, Google) or alternatives to EMR (Hortonworks HDP) can be used instead.

Finally, I implement the logic with MapReduce in Java thanks to libraries such as warc-hadoop which deals with the low-level access to WARC files. If you need to process CommonCrawl with existing frameworks and libraries such as Apache UIMA, Tika or GATE, our good old open source project Behemoth could help as it can ingest WARCs too!


As we’ve seen, CommonCrawl is an awesome resource and should be the first thing you try before embarking on web scale crawling (although if you must, DigitalPebble would be happy to help). It is large, it is free, it is relatively easy to process and a lot of effort has been put into making your life easier.

Web data are big, messy and often don’t give the results you expect. Processing the CommonCrawl dataset is a great way of checking your assumptions at a fraction of the cost of a web scale crawl. It also saves you time, as the fetch politeness has been done for you but on the minus side, you will be able to process only content allowed by robots.txt directives as CommonCrawl’s crawler is polite (but then yours should be too).

I hope you will give CommonCrawl a try and if you find it useful, you can donate to the project.

Thursday 23 March 2017

What’s new in StormCrawler 1.4

StormCrawler 1.4 has just been released! As usual, all users are advised to upgrade to this version as it fixes some bugs and contains quite a few new functionalities.

Core dependencies upgrades

  • Httpclient 4.5.3
  • Storm 1.0.3 #437

Core module

  • JSoupParser does not dedup outlinks properly, #375
  • Custom schedule based on metadata for non-success pages, #386
  • Adaptive fetch scheduler #407
  • Sitemap: increased default offset for guessing + made it configurable  #409
  • Added URLFilterBolt + use it in ESSeedInjector #421
  • URLStreamGrouping 425
  • Better handling of redirections for HTTP robots #4372d16
  • HTTP Proxy over Basic Authentication #432
  • Improved metrics for status updater cache (hits and misses) #434
  • File protocol implementation #436
  • Added CollectionMetrics (used in ES MetricsConsumer + ES Spout, see below) #7d35acb


  • Added code for caching and retrieving content from AWS S3 #e16b66ef


  • Basic upgrade to Solr 6.4.1
  • Use ConcurrentUpdateSolrClient; #183


  • Various changes to StatusUpdaterBolt
    Fixed bugs introduced in 1.3 (use of SHA ID), synchronisation issues, better logging, optimisation of docs sent and more robust handling of tuples waiting to be acked (#426). The most important change is a bug fix whereby the cache was never hit (#442) which had a large impact on performance.
  • Simplified README + removed bigjar profile from pom #414
  • Provide basic mapping for doc index #433
  • Simple Grafana dashboard for SC metrics, #380
  • Generate metrics about status counts, #389
  • Spouts report time taken by queries using CollectionMetric, #439 - as illustrated below
Spout query times displayed by Grafana
(illustrating the impact of SamplerAggregationSpout on a large status index )

Coming next?

As usual, it is not clear what the next release will contain but hopefully, we'll switch to Elasticsearch 5 (you can already take it from the branch es5.3) and provide resources for Selenium (see branch jBrowserDriver). As I pointed out in my previous post, getting early feedback on work in progress is a great way of contributing to the project.

We'll probably also upgrade to the next release of crawler-commons, which will have a brand new SAX-based Sitemap parser. We might move to one of the next releases of Apache Storm, where a recent contribution I made will make it possible to use Elasticsearch 5. Also, some of our StormCrawler code has been donated to Storm, which is great!

In the meantime and as usual, thanks to all contributors and users and happy crawling!

PS: I will be running a workshop in Berlin next month about StormCrawler, Storm in general and Elasticsearch

Friday 17 March 2017

Contribute to an open source project beyond code

I was recently contacted by someone who liked StormCrawler, wanted to contribute to it and asked me how to do so. While most contributions to open source projects take the form of code to either fix bugs or add new functionalities, there are various other ways in which people can contribute.

Here is what I replied to him, and while the examples below are about StormCrawler, the same ideas can apply to pretty much any open source project.

  • Spread the word: if you use StormCrawler, why not blog/tweet about it and get listed on the powered by page? The more people see that it is used, the more confident they become in adopting it. If you are not too shy: why not give a short presentation at a local tech meetup or a bigger conference?
  • Help with the documentation or write tutorials: we have WIKI pages and various instructions on the site - going through those would be a good way of learning about Apache Storm and StormCrawler while at the same time make a useful contribution.
  • Find bugs and possible improvements: run the code, benchmark it, look at the logs for unexpected things. Just play and see! If something is not clear, then the docs can be improved (see previous point).
  • Test things in branches / PRs: for instance, I started work on jBrowserDriver and Elasticsearch 5. Giving new functionalities an early try is fab.
  • Help others:  you have used StormCrawler a bit? Join the mailing list or follow StackOverflow and help newcomers overcome the hurdles as you did.
  • Donate resources: your company has one or more servers they are not using (unlikely but who knows)? You have AWS credits and don't know what to do with them? We can always do with test machines.
Any of those forms of contributions is valuable! Writing code is good but that's just one part of making a project successful.

PS: if you are wondering what happened with that prospective contributor, he's taken one of the open issues and doing great work on it! 

Tuesday 10 January 2017

What's new in StormCrawler 1.3

StormCrawler 1.3 has just been released! As usual, all users are advised to upgrade to this version as it fixes some bugs and contains quite a few new functionalities and improved performance. 

Dependencies upgrades

  • Jsoup 1.10.1
  • Crawler-Commons 0.7
  • RomeTools to 1.7.0
  • ICU4J 58.2

Core module

  • Hardcoded limit to the max # connections allowed by protocol #388
  • LangID module #364
  • JsoupParserBolt can use first N bytes for charset detection (or not at all) #391
  • SimpleFetcherBolt uses allowRedir from super class #394 (bugfix)
  • URLNormalizer : Decode non-standard percent encoding prior to re-encoding
  • MaxDepthFilter defaults to -1, 0 removes all outlinks, can set a custom max depth per URL with max.depth. Implements #399 and #400

The latter breaks compatibility with the previous versions: 0 was used to deactivate the filtering by depth, whereas now it is used to prevent any outlinks from being processed. Please change your config to -1 if you want to deactivate the filtering.


  • Flux for crawl and injection topologies #372
  • Use min delay for all types of Spouts #370
  • Remove Node client #377
  • ESSpout deals with deep paging before building query
  • Topology status updater triaged by URL to hit cache
  • Settings done via configuration #376
  • Add plugin to the clients via configuration #378
  • Spouts: load results with a non-blocking call #371
  • Concurrent requests in config #382
  • StatusUpdaterBolt - do not add URL already in buffer for ES if status is DISCOVERED
  • Allow fieldNameForRoutingKey to be outside metadata and use a different key for spouts #384
  • Use SHA256 as doc_id #385
  • Separate Kibana schema for status and metrics + put all schemas in a separate folder
  • Improvements to ES_IndexInit
  • ES crawl topology uses FetcherBolt
Please note that the cluster name is now defined alongside the other settings:
  es.status.settings: "elasticsearch"
One of the benefits of #376 and  #378 is that you can now use StormCrawler with Elastic Cloud protected with Shield.

We are fast approaching our 1.000th commit! Thanks to all users and contributors for their help with StormCrawler. Happy crawling!

PS: I will be running a 1-day workshop in Berlin on the 2nd of February. Announcements will be made on our Twitter account

Tuesday 3 January 2017

The Battle of the Crawlers : Apache Nutch vs StormCrawler

Happy New Year everyone!

For this first blog post of 2017, we'll compare the performance of StormCrawler and Apache Nutch. As you probably know, these are open source solutions for distributed web-crawling and we provided an overview last year of both as well as a performance comparison when crawling a single website.

StormCrawler has been steadily gaining in popularity over the last 18 months and a frequent question asked by prospective users is how fast it is compared to Nutch. Last year's blog post provided some insights into this but now we'll go one step further by crawling not a single website, but a thousand. The benchmark will still be on a single server though but will cover multi-million pages.

Disclaimer: I am a committer on Apache Nutch and the author of StormCrawler.

Meet the contestants

Please have a look at our previous blog post for a more detailed description of both projects. This Q and A should also be useful. 

Apache Nutch is a well-established web crawler based on Apache Hadoop. As such, it operated by batches with the various aspects of web crawling done as separate steps (e.g. generate a list of URLs to fetch, fetch, parse the web pages and update its data structures.

In this benchmark, we'll use the 1.x version of Nutch. There is a 2.x branch but as we saw in a previous benchmark, it is a lot slower. It also lacks some of the functionalities of 1.x and is not actively maintained.

StormCrawler, on the other hand, is based on Apache Storm, a distributed stream processing platform. All the web crawling operations are done continuously and at the same time.

What we can assume (and observed previously) is that StormCrawler should be more efficient as Nutch does not fetch web pages continuously, but only as one of the various batch steps. On top of that, some of its operations - mainly the ones that deal with the crawldb, the datastructure used by Nutch - take increasingly longer as the size of the crawl grows.

The battleground

We ran the benchmark on a dedicated server provided by OVH with the following specs :

Intel  Xeon E5 E5 1630v3 4c/8t  3,7 / 3,8 GHz
64 GB of RAM DDR4 ECC 2133 MHz
2x480GB RAID 0 SSD

Ubuntu 16.10 server

We installed the following software

Apache Storm 1.0.2
Elasticsearch 2.4
Kibana 4.6.3

StormCrawler 1.3-SNAPSHOT

Hadoop 2.7.3
Apache Nutch 1.13-SNAPSHOT

Finally, the resources and configurations for the benchmark can be found on the first release of

We followed the recommendations from 
for the configuration of Elasticsearch on SSD and gave it 10GB RAM to run on.

Apache Nutch 

The configuration for Nutch can be found in the GitHub repo under the nutch directory. This should allow you to reproduce the benchmarks if you wished to do so.

The main changes to the crawl script, apart from the addition of a contribution I recently made to Nutch, was to : 

  • set the number of fetch threads to 500
  • change the max size of the fetchlist to 50,000,000
  • use 4 reducer tasks
  • remove the link inversion and dedup steps

The latter was done in order to keep the crawl to a minimum. We left the setting for the limitation of fetch time to 3 hours. The aim of this was to avoid long tails in the fetching step, where the process is busy fetching from only a handful of slow servers. 

In order to optimise the crawl, we limited the number of URLs per hostname in the fetchlist to 100, which guarantees a good distribution of URLs and again, prevents the long tail phenomenon, which is commonly observed with Nutch. We also tried to avoid the conundrum whereby setting too low a duration for the fetching step requires more crawl iterations, meaning that more generate and update steps are necessary.

Note: we initially intended to index the documents into Elasticsearch, however, this step proved unreliable and caused errors with Hadoop. We ended up deactivating the indexing step from the script, which should benefit Nutch when comparing to StormCrawler.

We ran 10 crawl iterations between 2016.12.16 11:18:37 CET and 2016.12.17 19:29:30 CET, the breakdown of times per step is as follows : 

Iteration #StepsTime

What you can observe is that the generate and update steps do take an increasingly longer time, as mentioned above.

The final stats from the final update step were :


which gives us a total of 11,653,410 URLs processed (fetch + gone + redirs) in a total time of 1930 minutes.

On average, Nutch fetched 6,038 URLs per minute.

The graph below shows the bandwidth usage of the server when the Nutch crawl was running.

Network graph of Nutch crawl

This is a good illustration of the batch nature of Nutch, where the fetching is only one part of the whole process.

Let's now see how StormCrawler fared in a similar situation.


StormCrawler can use different backends for storing the status of the URLs (i.e. which is what the crawldb does in Nutch). For this benchmark, we used the Elasticsearch module of StormCrawler as it is the most commonly used. This means that we won't just be storing the content of the webpages to Elasticsearch, we'll also be using it to store the status of the URLs as well as displaying metrics about the crawl with Kibana.

We ran the crawl for over 2 and a half days and got the following values in the status index

which means a total of 38,943,845 webpages processed over 3977minutes, i.e. an average of 9792.26 pages per min.

The network graph looked like this:

Network graph of StormCrawler crawl

which, apart from an unexplained and possibly unrelated spike on Christmas day, shows a pretty solid use of the bandwidth. Whereas Nutch was often around the 50M mark, StormCrawler is lower but constant.

The metrics stored in Elasticsearch and displayed with Kibana gave a similar impression:

StormCrawler metrics displayed with Kibana
Interestingly, the Storm UI indicated that the bottleneck of the pipeline was the update step, which is not unusual given the 'write-heavy' nature of StormCrawler.


This benchmark as set out above shows that StormCrawler is 60% more efficient than Apache Nutch. We also found StormCrawler to run more reliably than Nutch but this could be due to a misconfiguration of Apache Hadoop on the test server. We had to omit the indexing step from the Nutch crawl script because of reliability issues, whereas the StormCrawler topology did index the documents successfully. This would have added to the processing time of Nutch.

The main explanation lies in the design of the crawlers: Nutch achieves greater spikes in the fetching step but does not fetching continuously as StormCrawler does. I had compared Nutch to a sumo and StormCrawler to a ninja previously but it seems that the tortoise and hare parable would be just as appropriate.

It is important to bear in mind that the raw performance of the crawlers is just one aspect of an overall comparison. One should also consider the frequency of releases and contributions as well as more subjective aspects such as the ease of use and versatility. There is also of course the question of the functionalities provided. To be fair to Nutch, it currently does thing that StormCrawler does not yet support such as document deduplication and scoring. On the other hand, StormCrawler too has a few aces up its sleeve with Xpath extraction, sitemap processing and live monitoring with Kibana.

As often said in similar situations: “your mileage might vary”. The figures given here depend on the particular seed list and hardware, you might get different results on your specific use case. The resources and configurations of the benchmark being publicly available, you can reproduce it and extend it as you wish.

Hopefully we’ll run more benchmarks in the future. These could cover larger scale crawling in fully distributed mode and/or comparing different backends for StormCrawler (e.g. Redis+RabbitMQ vs Elasticsearch).

Happy crawling!