Currently I am experimenting with ways to reduce server uptime in a manner that does not involve clearing props out of the database.
This involves having a set of global props that are always loaded while also loading and unloading player props. This will reduce the entity count and hopefully result in an uptime increase.
Your furniture might not load, may act odd, or will do odd stuff as I look at the errors and fix it. Global props seem to have been loading without problem. It's mainly just player props that aren't loading right.
Admins may look at the main door of an apartment and use the command: /loaddoorfurniture to load the props associated with that door. /unloaddoorfurniture can also be used to unload furniture linked to a door. The plugin might tell you that the props for a door are already loaded. If it does, just do /unloaddoorfurniture and then /loaddoorfurniture. The props should then spawn.
Also with the help of Syle, I have identified the specific code paths leading to the "memory leak" issue. Now I know exactly how to fix it. This will take me a day or two. The above two fixes should bring server uptime from the 5-12hr range to the 1-2 day range.
This involves having a set of global props that are always loaded while also loading and unloading player props. This will reduce the entity count and hopefully result in an uptime increase.
Your furniture might not load, may act odd, or will do odd stuff as I look at the errors and fix it. Global props seem to have been loading without problem. It's mainly just player props that aren't loading right.
Admins may look at the main door of an apartment and use the command: /loaddoorfurniture to load the props associated with that door. /unloaddoorfurniture can also be used to unload furniture linked to a door. The plugin might tell you that the props for a door are already loaded. If it does, just do /unloaddoorfurniture and then /loaddoorfurniture. The props should then spawn.
Also with the help of Syle, I have identified the specific code paths leading to the "memory leak" issue. Now I know exactly how to fix it. This will take me a day or two. The above two fixes should bring server uptime from the 5-12hr range to the 1-2 day range.