User Blogs

I recently started playing Fallout 4 again, because I own it already and it runs well on my potato, even under Linux. The last time I played I had dual nvidia GTX 950 cards in SLI under Windows 7, and now I have a single 1070 under Linux. The performance is as much better as you would hope. The game is just about as reliable on Linux as it is on Windows, which is not very. However, my current configuration lets the game start up pretty quickly, and liberal use of the quicksave feature keeps me from losing too much progress in the inevitable explosions.

composer config repositories.asset-packagist composer https://asset-packagist.org
composer config --json --merge extra.installer-types '["bower-asset", "npm-asset"]'
composer config --json --merge 'extra.installer-paths.web/libraries/{$name}' '["type:bower-asset", "type:npm-asset"]'

(From https://www.drupal.org/project/drupal/issues/3198245#comment-13997741)

I am using the AUTOMATIC1111 stable-diffusion distribution, and you can install modules to that. Some modules can be installed and updated from inside of the program, while others cannot. Those other modules have to be updated by doing a git pull. This is not a big job, but it can be made even smaller with a very small shell script.

I wanted a nice way to simply see whatever statistics I could gather from my GPU — ye olde PNY XLR8 1070 slightly-pre-overclocked model, which I bought used and which has been surprisingly solid so far — and for some reason the information I want to see isn't in the default output from the nvidia-smi command included with the nvidia driver, and I don't get anything in lm-sensors. Noodling around a bit led me to the idea of using the --query-gpu option, which lets you pull any specific information you want, as long as you are willing to deal with it in CSV format. I was, and here we are.

Without getting [too] technical (and probably operating "above my pay grade") the ZFS ZIL is a place where synchronous writes are done, and you can move this to a separate zpool, which is called a SLOG. One SLOG can store the ZIL for multiple zpools. While most writes by most applications are asynchronous, some are not, and file metadata writes are synchronous. Putting the ZIL for a HDD (or even an array) on a SSD lets your zpool provide the IOPS performance of your SLOG, while also avoiding fragmentation and reducing seeks through ordered writes.