ReSeeD supports large datasets up to multiple terabytes. However, uploads through the web browser are limited to a small fraction of this size. Importing larger amounts of data requires administrative access to ReSeeD, which is not available through self-service.
Our team will gladly assist you with importing large datasets. A consultation with our staff is mandatory before an assisted import. Please let us know if you are interested in using this service (reseed@ruhr-uni-bochum.de).
The following information provides an overview of the process and the requirements for your data and metadata.
Please keep in mind that assisted import is a time-consuming process. It involves a mandatory consultation, data preparation on your side, data ingest on our side, metadata editing on your side, and, if desired, additional workflows for archiving or publishing the dataset. Please communicate any deadlines to us as early as possible.
Before our team can assist you with the import, you need to prepare your data.
Please carefully read our submission guideline, especially the information about file name restrictions.
Please keep the number of files below 100. Navigating and managing larger numbers of files in the web interface becomes difficult, and performance may decrease. You may pack your data into containers or archives. We recommend to limit the size of each individual file to no more than 50 GB. This size still allows users to download individual files on most regular computers.
Please note that data compression, even if lossless, can make long-term data preservation more difficult. ReSeeD will not process or unpack containers or archives on the server side.
The directory structure of your data will be preserved in ReSeeD during ingest, but it cannot be changed later. This requires careful planning of your directory layout. When users download the entire dataset using Download All, they receive a shell script that recreates the directory structure on their local file system.
ReSeeD currently does not offer a tree view for navigating directory structures in the user interface. Instead, the list of items attached to a dataset shows the relative file paths of each file (e.g., day01/measurementA/results.csv).
Individual files can be replaced or deleted, but their path within the dataset remains unchanged. Manually adding files is possible, but these files can only be added at the top level of the dataset's directory structure (e.g., a README.md file).
A detailed README file as well as rich metadata are fundamental to making large amounts of data findable and reusable. We also encourage you to provide additional forms of data documentation, such as data dictionaries or codebooks.
For large datasets, we strongly recommend providing example files of smaller size. These allow other researchers to get a better understanding of your dataset without having to download large amounts of data.
An example of an imported dataset (3 TB) is available at: Bocchini, G. M., Roth, M. P., and Harrington, R. M.: Earthquake Catalog and Continuous Waveforms From a Two-week Distributed Acoustic Sensing Experiment on Kefalonia Island, Greece, ReSeeD [data set], https://doi.org/10.60517/cv43p1601, 2025. The dataset is organized by date, with each day containing four .tar archives corresponding to six-hour intervals:
├── 2024-08-02
│ ├── 00_06.tar
│ ├── 06_12.tar
│ ├── 12_18.tar
│ └── 18_24.tar
├── 2024-08-03
│ ├── 00_06.tar
│ ├── 06_12.tar
│ ├── 12_18.tar
│ └── 18_24.tar
└── 2024-08-04
...
For ingest, we need a descriptive and individual title for your dataset so that it can be clearly identified in ReSeeD.
No other metadata is required at the time of data ingest. You can add the remaining metadata later using the edit form in the web interface. Our team will fill any required metadata fields with placeholders for you to replace.
During ingest, data will be read from a bucket on RUB's CEPH object storage. This system is S3-compliant and can be accessed using rclone or similiar tools.
Following the consultation with our staff, we will provide you with credentials for accessing a temporary upload bucket.