You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Bug Details
Put a 7z file with 1000 images into content directory, and open that file in the website to start the page_thumbnails task, and it will use about 2 hours with 100% cpu usage (1 core) to generate 1000 thumbnails .
The page thumbnail mechanism has been reworked to run sequentially on a single job as of the current version in order to avoid spamming all cores and filling the runner with 1 job per page when a lot of pages need to be thumbnailed.
I'd have to reintroduce some small multithreading within the single job to improve this.
However, there is caching in place, so even if the job is halted or restarted, it'll pick back up wherever it stopped creating thumbnails.
LRR Version and OS
v0.9.21
Docker
Bug Details
Put a 7z file with 1000 images into content directory, and open that file in the website to start the page_thumbnails task, and it will use about 2 hours with 100% cpu usage (1 core) to generate 1000 thumbnails .
CPU: i5-1340P
RAM: 16GB DDR4 3200Mhz
Disk: NVMe SSD
Matching Logs
https://github.com/Difegue/LANraragi/blob/v.0.9.21/lib/LANraragi/Utils/Minion.pm#L75
https://github.com/Difegue/LANraragi/blob/v.0.9.21/lib/LANraragi/Utils/Archive.pm#L155
Maybe it need a cache for
extract_thumbnail
?The text was updated successfully, but these errors were encountered: