- MCP server with stdio transport for local use - Search episodes, transcripts, hosts, and series - 4,511 episodes with metadata and transcripts - Data loader with in-memory JSON storage 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
53 lines
3.8 KiB
Plaintext
53 lines
3.8 KiB
Plaintext
Episode: 1540
|
|
Title: HPR1540: The Journaling File System
|
|
Source: https://hub.hackerpublicradio.org/ccdn.php?filename=/eps/hpr1540/hpr1540.mp3
|
|
Transcribed: 2025-10-18 04:47:51
|
|
|
|
---
|
|
|
|
You
|
|
Okay guys, hey welcome and I hope all are having a great day.
|
|
I am continuing my thing on file systems, different ones that I've used, ones that I've found
|
|
and things that are going on.
|
|
Work we've been having a lot of talk about different file systems and what's going on
|
|
and I do a lot of ERP stuff and so much, much talk about BTFRS and XFS and different things
|
|
and I just thought I'd do a little research and maybe share some of that research with you.
|
|
So we covered XT, the original thing that came in Linux kernel 9.6 and we talked about XT2
|
|
and that you know it had originally two terabytes and then was upgraded to 32 terabytes
|
|
and that it had lots of things and so the way that you fixed most of the errors in here,
|
|
particularly the in node error that XT2 had that was so wrong but that I never had a problem with
|
|
but if you read online that there's lots of people that had lots of problems with it.
|
|
So the way to get rid of that is with something called a journaling file system.
|
|
And journaling file systems provide a new level of safety to the system
|
|
instead of writing directly to the storage device when updating the in node table,
|
|
journaling file systems write file changes to a temporary file called the journal first.
|
|
And after the data is successfully written to the storage device, in node table, the journal entry is deleted.
|
|
If the system should crash or suffer a power added before the data can be written to the storage device,
|
|
the journaling file system just needs to read through the journal file and process any uncommended data leftover.
|
|
There are three different methods of journaling commonly used each with different levels of protection.
|
|
The first is a thing called data mode and in data mode both the in node and the file data are journaled.
|
|
This offers a low risk of losing data but very poor performance.
|
|
And then they have something called the ordered mode.
|
|
Only the in node data is written to the journal but not removed until after the file is successfully written.
|
|
It's a good compromise between performance and safety.
|
|
Then there's something called the write back mode.
|
|
Only in node data is written to the journal.
|
|
The control over when the file data is written.
|
|
It's a higher risk of losing data but is still better than not using journaling.
|
|
The limitation, the data mode journaling method by is by far the safest for protecting data but is also the slowest.
|
|
All of the data written to the storage device must be written twice, once to the journal and then again to the actual storage device.
|
|
This can cause poor performance especially for systems that do a lot of writing.
|
|
Over the years a few different journaling file systems have appeared and we'll talk about a lot of those later.
|
|
Okay, I hope that you've enjoyed this installment of the different file systems.
|
|
If you want to contact me at JWP5 at hotmail.com, thank you very much.
|
|
You have been listening to Hacker Public Radio where Hacker Public Radio does our.
|
|
We are a community podcast network that releases shows every weekday Monday through Friday.
|
|
Today's show, like all our shows, was contributed by a HBR listener like yourself.
|
|
If you ever consider recording a podcast, then visit our website to find out how easy it really is.
|
|
Hacker Public Radio was founded by the digital dog pound and the infonomicum computer globe.
|
|
HBR is funded by the binary revolution at binref.com.
|
|
All binref projects are proudly sponsored by LUNAR pages.
|
|
From shared hosting to custom private clouds, go to LUNAR pages.com for all your hosting needs.
|
|
Unless otherwise stasis, today's show is released under a creative comments, attribution, share a life, free those own license.
|
|
Hacker Public Radio is funded by the digital dog pound and the infonomicum computer globe.
|