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
As the memory usage of the chainbase db is growing rapidly, we would like for Spring to be able to work well even when the available RAM is less that the amount of memory allocated in the chainbase db.
In order to have a better understanding of the current memory usage patterns in Spring, we propose to implement a graphical visualization of the memory as a square map of pages, whith a color coding indicating the percentage of the page that is allocated (bright red would be 100% allocated, bright green would be 0% allocated).
If the display window is 1024 x 1024 pixels, and we want to be able to display the allocation pattern in 512 gb of memory, each pixel would represent 512 kb of memory, or 128 4k pages. The visualization will provide a zoom functionality, so the allocation of individual pages can be visualized.
Also, we plan to investigate whether it would be possible to visualize the access pattern into the memory from Spring as it is running. It may provide insights about possible reorganization of the storage patterns, so that frequently accessed memory has better locality.
The text was updated successfully, but these errors were encountered:
As the memory usage of the chainbase db is growing rapidly, we would like for Spring to be able to work well even when the available RAM is less that the amount of memory allocated in the chainbase db.
In order to have a better understanding of the current memory usage patterns in Spring, we propose to implement a graphical visualization of the memory as a square map of pages, whith a color coding indicating the percentage of the page that is allocated (bright red would be 100% allocated, bright green would be 0% allocated).
If the display window is 1024 x 1024 pixels, and we want to be able to display the allocation pattern in 512 gb of memory, each pixel would represent 512 kb of memory, or 128 4k pages. The visualization will provide a zoom functionality, so the allocation of individual pages can be visualized.
Also, we plan to investigate whether it would be possible to visualize the access pattern into the memory from Spring as it is running. It may provide insights about possible reorganization of the storage patterns, so that frequently accessed memory has better locality.
The text was updated successfully, but these errors were encountered: