When N is bigger, the ratio will be smaller. If N >= 1, the ratio
will be [15/1024, 4/1024), for this the ratio 15 : 1024 will be
enough. But maybe some iscsi cmds has no datas, N == 0. So the ratio
should be bigger.
For now we will increase the data area size to 1G, and the cmd area
size to 128M. The tcmu-runner should mmap() about (128M + 1G) when
running and the TCMU will dynamically grows the data area from 0 to
max 1G size.
Cool. This is a good approach for an initial patch but this raises
concerns about efficiently managing kernel memory usage -- the data area
grows but never shrinks, and total possible usage increases per
backstore. (What if there are 1000?) Any ideas how we could also improve
these aspects of the design? (Global TCMU data area usage limit?)
Two ways in my mind:
The first:
How about by setting a threshold cmd(SHRINK cmd), something likes
the PAD cmd, to tell the userspace runner try to shrink the memories?
When the runner get the SHRINK cmd, it will try to remmap uio0's ring
buffer(?). Then the kernel will get chance to shrink the memories....
The second:
Try to extern the data area by using /dev/uio1, we could remmap the
uio1 device when need, so it will be easy to get a chance to shrink the
memories in uio1.
Maybe these are a little ugly, are there other more effective ways ?
Thanks,
BRs
Xiubo
The cmd area memory will be allocated through vmalloc(), and the data
area's blocks will be allocated individually later when needed.
