feynon / crawler

Explore the web in parallel on thousands of machines

Crawler

Explore the web in parallel on thousands of machines

TODO

• Productionize master & worker with app framework
• Configure continuous delivery to bootstrap productivity
• Health check for workers & master
• Abstract out worker scheduling to support custom algorithms
• Assign job to new worker when previously assigned worker failed
• Support checkpoint for master
• Auto recover system when the master is back online
• Support pluggable worker script
• Load worker script from network file system
• TLS (Transport Layer Security)
• Create CLI to trigger crawling for a certain site
• Support Docker Swam
• Support k8s

Getting Started

cd bin

Start Master

go run master.go 8080

Output

Master started at 8080

Start Workers

Run the following command at different terminals:

go run worker.go 8081 localhost 8080
go run worker.go 8082 localhost 8080
go run worker.go 8083 localhost 8080

Output

At the master side:

Worker registed: ID(0) IP(localhost) PORT(8081) SECRET(encrypted)
Worker registed: ID(1) IP(localhost) PORT(8082) SECRET(encrypted)
Worker registed: ID(2) IP(localhost) PORT(8083) SECRET(encrypted)

At the work side:

Registered with master: id(0)
Registered with master: id(1)
Registered with master: id(2)

Try It Out

Send the following gRPC calls to the master:

{
 "url": "https://leetcode.com"
}

Output

On the master side:

Start exploring https://leetcode.com
https://leetcode.com
/support/
/jobs/
/bugbounty/
/terms/
/privacy/
/region/
mailto:[email protected]?subject=Billing%20Issue&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
mailto:[email protected]?subject=General%20Support&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
mailto:[email protected]?subject=Other%20Inquiries&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
Finish exploring https://leetcode.com

On the worker side:

// Worker 0

Start extracting links from https://leetcode.com
/support/
/jobs/
/bugbounty/
/terms/
/privacy/
/region/
mailto:[email protected]?subject=Billing%20Issue&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
mailto:[email protected]?subject=General%20Support&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
mailto:[email protected]?subject=Other%20Inquiries&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
Start extracting links from /support/
Start extracting links from /bugbounty/
Start extracting links from /region/
Start extracting links from /privacy/

// Worker 1
Start extracting links from /jobs/
Start extracting links from mailto:[email protected]?subject=General%20Support&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A

// Worker 2
Start extracting links from mailto:[email protected]?subject=Other%20Inquiries&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
Start extracting links from mailto:[email protected]?subject=Billing%20Issue&body=Name:%0D%0A%0D%0AUsername:%0D%0A%0D%0AMessage:%0D%0A%0D%0A
Start extracting links from /terms/

Author

• Yang Liu - Initial work - byliuyang
• Vinod Krishnan - Incremental improvements - vtkrishn

License

This project is maintained under MIT license.