Dump structure

Every time you make a dump using CleverAnalytics Shell, a dump directory structure is created. Every dump is uniquely identified by its dump ID - a time stamp of when the dump was created. Format of the dump ID is "yyyy-MM-dd_HH-mm-ss".

All dumped metadata objects are saved to the file system in a wrapper, which contains detailed metadata about the object itself.

In addition, dump metadata (or "metadata about the metadata") file is created. It is used to store useful information, like dump time, or the original project ID. It is also used to store the dump content and to track local file changes.

Dump structure

Dump directory structure is created either in:

1. the default directory you specified in the config file (recommended)
2. working directory, e.g. the directory you run Shell from
3. the directory you specify as a value of --directory parameter of the dumpProject command

Let's assume that in this case, /var/local/metadata is our default directory. The dump structure is following:

/var/local/metadata/{projectId}/{dumpId}/

So, in practice:

/var/local/metadata/q1zdp9d0ao78rdv5/2017-08-28_20-17-51/

If we list the contents of this dump directory, we see data and metadata directories, and the dumpMetadata.json file (described below).

./
../
data/
dumpMetadata.json
metadata/
metadata/dashboards/
metadata/dataPermissions/
metadata/datasets/
metadata/exports/
metadata/indicatorDrills/
metadata/indicators/
metadata/markers/
metadata/markerSelectors/
metadata/metrics/
metadata/projectSettings/
metadata/views/

The data folder contains .csv files. The metadata folder contains subdirectories of the metadata object types. Each dumped file is wrapped in a dump object wrapper.

Dump object wrapper

Simple wrapper, which contains metadata about the dumped object.

Dump metadata file

The dumpMetadata.json file is located the dump directory, one level above the data and metadata directories. It contains metadata about the dump itself, and list of its contents.