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
We have different requirements for limits between community and commercial -- namely we allow a lot more build time by default on commercial. The most explicit version of this would be using the same logic to check memory, but pushed into a @property setting in community/commercial.
Because of the removal of all docker limit settings between community/commercial/ops, the default build time on commercial has dropped.
Also, there is a lot of information squirreled away in the DMI information: sudo find /sys/devices/virtual/dmi/id/ -type f | sudo xargs -n1 -t cat
This could be of use if we want to detect default settings only when running in production perhaps. Having explicit limits on readthedocsinc.settings.base might be enough.
Now it's dynamically calculated based on RAM and using a coefficient to calculate the time. There were some adjustment to that logic over time and I think that it's working as we want now: higher memory and time on commercial than on community.
We have different requirements for limits between community and commercial -- namely we allow a lot more build time by default on commercial. The most explicit version of this would be using the same logic to check memory, but pushed into a
@property
setting in community/commercial.Because of the removal of all docker limit settings between community/commercial/ops, the default build time on commercial has dropped.
Also, there is a lot of information squirreled away in the DMI information:
sudo find /sys/devices/virtual/dmi/id/ -type f | sudo xargs -n1 -t cat
This could be of use if we want to detect default settings only when running in production perhaps. Having explicit limits on readthedocsinc.settings.base might be enough.
This was discovered in #6982
The text was updated successfully, but these errors were encountered: