I finally got the opportunity to examine the OP video thoroughly after downloading it on my computer. First of all, the contrail in the beginning is not a short one that evaporates quickly. It is a time lapse movie made with very wide 'fish eye' lenses, suggesting that the trail actually was much longer and lasted quite a while. In fact, its length and duration can be deduced from the video and related information.
The camera faces South; at the solar noon the Sun is right above the building in the middle. The Sun elevation at this point on the date (28-Jan-2015) was 34°. There are more time lapse videos with the same set up on the author's channel. Some of them captured 22° halo, for example,
the one taken on 11-Feb-2015:

The halo provides the 'scale bar' for angular dimensions. In particular, the angular width of the tall building in the middle is also about 22°. It takes approximately 1.5 hours for the Sun moving over the building from the left side to the right. In the OP video, the Sun does it in about 17 seconds. It means that the time has been speeded up about 300x, or about 5 minutes of real time per one second of the video. As the movie rate is 24 fps, it corresponds to 12.5 seconds per frame.
It takes 3 to 4 frames for the contrail moving along by its full length, meaning the trail persisted for 40-50 seconds after it was formed. The screenshot below shows the contrail 4 frames = 50 seconds after it first appeared from behind the building:

I've used the Sun elevation at the solar noon (34° on 1-28-15, 38° on 11-2-15) as a reference for the determination of the angular positions of the contrail ends. The bottom end is at about 24° above the horizon, the top one is at 39° above the horizon and about 22° further to the right. In the assumption of a level flight at the altitude of 10 km, the calculated contrail length is about 12 km. Thus the estimated plane speed is about (12 km/50 sec)x3600 = 860 km/h, which is in the range of the cruise speed values of commercial jets.
The contrail lifetime is significantly longer than the lifetimes of short contrails (<10 sec), as seen in this real-time video:
but it is comparable with the lifetime of
"hybrid contrails", limited by the Crow instability:
This analysis supports my suggestion in the previous post (
#25), that the trail in the beginning of the OP movie was (a pair of) the hybrid contrail(s). As the above link show, at certain atmospheric conditions the hybrid contrails co-occur with ordinary exhaust contrails that may have different initial densities. The "reappearing" trail was most likely such an exhaust trail, the initial density of which was to weak to be picked up by the wide angle camera lenses. However, enhancing a screenshot from the video in full HD on a big screen suggests its presence where the initial, hybrid contrail has already dissipated:
