Grokking Concurrency in Python
0% completed
Memory in Worker Processes
A forked child starts with all of the parent's memory, and the kernel copies none of it at the moment of the fork. That memory stays shared only while neither process writes to it. In CPython, even reading an object writes to that object.
A service that forks eight workers to share one large index therefore often holds much more than one copy of it.
Copy-on-Write
Operating systems manage memory in pages. **A page is a fixed-size block of memory, 4 KB on most Linux systems, which the kernel maps into a process's address space as one unit
.....
.....
.....
Like the course? Get enrolled and start learning!
Reading Progress
0%