I took a look at the laser. It is probably the LD TEC (DTEC) failure.
As the temperature of the LD (DTMP) gradually deviated from 25degCish,
the DTEC voltage also went up from 2Vish to 2.1, 2.2...
When DTEC reaches 3V, it stopped lasing. This cools the diode a bit, and
it start lasing but repeat the above process.
I am not sure which of the head and controller has the issue.
The situation did not improve much by reducing the pumping current (ADJ: -15).
BTW, Turning on/off the noise eater did not change the situation.
I think the head/controller set should be sent out to JDSU and find how they will say.