-
|
Also raised this discussion on the k8s native helm chart repository, but no one seemed to notice. Upon testing, the AIO deployed in k8s is much faster than the native k8s helm chart. Maybe NC should consider migrating the build process for AIO images to the k8s native side, because the performance is much, much, much better, when both helm releases are installed on the same machine under the same network environment. The only thing that is slightly missing would be the more professional organization of helm values.yml, currently they feel like environment variables that mimic docker instead of following k8s structure. Thank you for making this releasing actually perform normally. Loading a page takes ~3 seconds whereas the k8s native one is around 6s on average, and could stretch to 12 seconds at times. |
Beta Was this translation helpful? Give feedback.
Replies: 2 comments 1 reply
-
Hi, thanks for this positive feedback! :)
Yes, this is indeed exactly what we are doing in order to keep all the different deployment methods in sync. See https://github.com/nextcloud/all-in-one/blob/main/nextcloud-aio-helm-chart/update-helm.sh. So we are not going to adjust their naming because this would mean to split the helm chart out of AIO which would potentially introduce issues because it would need to be kept in sync manually. |
Beta Was this translation helpful? Give feedback.
-
|
Makes sense. However, do you know why the AIO chart performs better than the native chart? What is your special sauce? This way I can make a PR on the native k8s images. |
Beta Was this translation helpful? Give feedback.
-
|
The aio images are tuned in various ways in order to make sure that it performs well |
Beta Was this translation helpful? Give feedback.
Hi, thanks for this positive feedback! :)
Yes, this is indeed exactly what we are doing in order to keep all the different deployment methods in sync. See https://github.com/nextcloud/all-in-one/blob/main/nextcloud…