shellpipe

shellpipe on Hacker News

First Steps into Memory Management in Python

Posted by shellpipe on Monday October 05, 2026 @04:01PM
from the tech.daniellbastos.com.br dept.

https://tech.daniellbastos.com.br/posts/python-memory-management/

( Read More... | 2 comments | on HN )

A 40ms Go garbage collector pause caused by swap

Posted by shellpipe on Sunday October 04, 2026 @09:11PM
from the frn.sh dept.

https://frn.sh/go-gc/

( Read More... | 189 comments | 13 points | on HN )

Re:Page Table Memory Consumption on Sunday October 04, 2026 @08:16PM (#49959354)

That's crazy! I appreciate you bringing this up. I will add this case as an interesting aside later.

Page Table Memory Consumption

Posted by shellpipe on Wednesday September 30, 2026 @09:55PM
from the frn.sh dept.

https://frn.sh/pagetables/

( Read More... | 13 comments | 3 points | on HN )

Simulating Airband Am Radios

Posted by shellpipe on Saturday September 26, 2026 @10:01PM
from the bitbashing.io dept.

https://bitbashing.io/am-radio.html

( Read More... | 5 comments | on HN )

A 40ms Go GC pause caused by swap

Posted by shellpipe on Sunday September 13, 2026 @11:11PM
from the frn.sh dept.

https://frn.sh/go-gc/

( Read More... | 1 comment | 6 points | on HN )

Swap, memory pressure and the Go GC

Posted by shellpipe on Tuesday September 08, 2026 @09:30PM
from the frn.sh dept.

https://frn.sh/swap/

( Read More... | no comments | on HN )

Re:Io_uring Without Readahead on Tuesday September 01, 2026 @05:13PM (#49528287)

TPC-H is the recommended approach in Turso's CONTRIBUTING.md - https://github.com/tursodatabase/turso/blob/main/CONTRIBUTIN...

Data Access Patterns That Makes Your CPU Angry

Posted by shellpipe on Saturday June 27, 2026 @07:32PM
from the blog.weineng.me dept.

https://blog.weineng.me/posts/slowest_add

( Read More... | no comments | on HN )

Re:Reading a cryptominer from its strace on Saturday June 13, 2026 @05:12PM (#48521532)

Not entirely sure, but my best guess is that the previous DevOps engineer was running a PostgreSQL 14 instance exposed to the internet with the password set to postgres. There's even an old CVE describing a remote code execution path for that kind of setup. Unfortunately, the PostgreSQL logs had been deleted, so I was never able to confirm it.