How can you easily create a phantom file when syncing cloud storage? (Yandex.Disk)

I discovered a great feature of Yandex-Disk, and I am sharing this wonderful information.



You can use this Yandex-Disk feature when you urgently need to create a lot of data, and it is inconvenient for you to load the I / O subsystem onto disks by copying files right now.



Initial requirements: You need to have a large amount of cloud storage on Yandex-Disk (do you need a lot of data?). It is advisable to synchronize several devices with this storage, for example, a desktop, laptop, and something else (you can have a working computer). The storage needs free space to accommodate a copy of the data.



Then you need to select a folder that contains the amount of data you need (for example, a lot. Say, a few gigabytes is a lot for Yandex-Disk). It is highly desirable that within this folder there are several levels of subfolders with files.



Then you just need to rename folders at several nesting levels on one of the synchronized devices.



The Yandex-Disk algorithm, developed by the programmers of the Yandex-Disk team, recognizes your desires, and during the next synchronization of devices with cloud storage, it will not perform a straightforward renaming of folders, but duplicate the folder tree with attached files, thus creating a lot of data for you, and at the same time creating a phantom file in one of the folders. Obviously, this will be done as an easter egg, greetings from the team of programmers from the Yandex-Disk team to your unconscious, which wanted to create a lot of data. The presence of a phantom file can be detected by looking at the file and folder counter in the Yandex-Disk web interface. One branch of the folder tree will have one more file than the other. This file is not observed on synchronized devices.



Yandex-Disk-team does not consider this feature to be a bug, denies the incorrect operation of this algorithm, and you can count on the feature to exist for a long time.



This experiment was carried out twice. Both times the feature worked with a bang, and a lot of data was created.



You can experiment, test this feature, and report your results.