Fedora 33 To Be Released Next Week

[S.Pam]

Thanks for that input - how would you activate that in Firefox?

This works with all sqlite dbs, unless the application is using ancient versions.

Always make a snapshot or backup of your files.

1. Close your application
2. Open the db: sqlite3 somedb.sqlite
3. Check current journal mode: PRAGMA journal_mode;
4. Enable WAL: PRAGMA journal_mode=WAL;
5. Compact db for good measure (not strictly needed): VACUUM;

Typing this on my phone so there might be mistakes... 🙃

[pal666]

Did i miss something?

yes. every lvm bulletpoint seriously sucks compared to btrfs

[pal666]

As ext4 outperforms btrfs i dont see your point.

but you are advertising ext4 on lvm. btrfs outperforms that abomination

[Girolamo_Cavazzoni]

but you are advertising ext4 on lvm. btrfs outperforms that abomination

Usually it's even worse: ext4 on LVM on LUKS. *eagerly waiting for native btrfs encryption*

[Mario Junior]

For me btrfs is just great. You want to create "mistake protection" (snapshot: ) of 500GB catalogue with photos before starting some risky operation? Copy-paste in Nautilus - done in seconds.

Copy-On-Write is excellent idea

With btrfs is much easier to "protect" system before updates - I had couple situations that after `dnf update` in Fedora - something stopped working (problem with hardware not supported in Linux - nVidia of course). With btrfs snapshot before update - "rollback" could be done almost immediately. If I'm not mistaken OpenSUSE is doing it already automaticly.

Opensuse has the best btrfs implementation of all linux distros.

[Space Heater]

Opensuse has the best btrfs implementation of all linux distros.

The kernel is what implements the filesystem, do you mean "integration"?

[S.Pam]

https://www.phoronix.com/scan.php?page=article&item=linux-58-filesystems

For those that think btrfs is a bad choice, those benchmarks show different.
Application start up would be much improved with BFQ scheduler though..

Also, if you enable WAL write ahead logging in sqlite, Btrfs becomes 3x faster!

[S.Pam]

I did some benchmarks with PTS SQLite-2.1.0 test with and without nocow as well as with and without Write-Ahead Logging.

The benchmarks show that by using Write-Ahead Logging, we can gain a 300% performance boost, compared to only about 25-30% using the nocow attribute. Not bad, huh?

I blogged about it here: https://wiki.tnonline.net/w/Blog/SQL...mance_on_Btrfs

EDIT: I did try to create a profile on openbechmarking.org, but the activation email doesn't work so I am stuck there Michael Can you help?