Comment 13 for bug 1776888

Hi Tone,

For point 1, we can close it, right.

For point 2, working well doesn't mean good code. I think the charm of the open source community is here. Good code can make it easier for others to read, easier to understand, and more convenient to maintain, isn't it?

For point 3, Can't the things released be changed? Where is the meaning of open source? Now the label logic is obviously more cumbersome to process, so why not use the previous logic?

For point 4, I am not just updating the documentation, I am re-defining the disk information of the three deployment methods in the logic of find_disks.py. I know the meaning of each partition of ceph. The third way is to provide a more free way to deploy ceph for users. It can be used for testing. It can also be used to customize the deployment that users want. I know that four partitions are deployed on a disk is meaningless, but users can test the deployment of the kolla with fewer disks, similar to deploying an osd with 4 loop devices, isn't it?

For point 5, you test no problem because you have not changed this place to bluestore:

```
OSD_INITIAL_WEIGHT=$(parted --script ${OSD_PARTITION} unit TB print | awk 'match($0, /^Disk.* (.*)TB/, a){printf("%.2f", a[1])}')
```
The partition information here should be block partitioned in the bluestore, isn't it?

Please take a look at my code flow and then compare it.

Thanks!