Yesterday, we aligned the bypassing path again.

At first, we checked where it was misaligned.
We found that the beam did not hit the center of the last PBS on PSL table to adjust the output power from the PSL room.
This means that the misalignment occurred somewhere on PSL table.

Then we aligned the bypassing path.
After we roughly aligned the pass so that the beam hit the center of the last PBS on PSL table, the fine alignment was done using QPDs for IMC WFS 
Specifically, we adjusted two mirrors, the flip mirror and the mirror just after second lens in bypassing path, and centered the beam on the two QPDs.
(Before this work, the main path beam on the QPDs were centered.)
As a result, the bypassing pass alignment was recovered.

In fact, we are still not sure why baypassing path was misaligned.
Thus, this problem could happen again.
We should check the bypaasing path alignment after a while.