Multiple datasets ----------------- Camelot offers the ability to manage multiple isolated collections of data (a "dataset") within a single running Camelot. There are other situations where you really want to keep data isolated from each other, for example, as a contractor using Camelot to conduct research for several clients. For this, Camelot offers datasets. Essentially you can think of each dataset as being a different Camelot database. Configuration ~~~~~~~~~~~~~ In the |administration_ui| you can add additional datasets and configure the location of the folders which comprise a dataset (more on this soon). You can also remove datasets. Note that the removal of datasets does **not** delete the data; removal means only that it will no longer be available through Camelot. Note that after changing dataset configurations, you must stop and then restart Camelot for those changes to take effect. .. figure:: screenshot/datasets.png :alt: .. |administration_ui| raw:: html administration interface Storage locations ================= These settings allows customisation of where Camelot stores its data. **Important**: Camelot will not move your data when customising any of these locations. When changing any of these locations, any existing data you wish to continue to use **must** be moved to this location manually. * **Media location**: where Camelot's images should be stored. **Note** that this should not be the same location as *you* store your images. Camelot will maintain its own copy of any images, folders and images stored in this location **should not** be modified. * **Database location**: the folder where Camelot will store its database. The Database folder **must not** be modified. If you have an existing database you wish to use, this should be set to the **parent** of the "Database" folder. * **FileStore location**: any files related to a survey can be uploaded to Camelot under the "survey files" menu. These files will be stored in this folder. Files and folders in this location **should not** be modified. * **Backup location**: Camelot will automatically back up its database before doing a database upgrade. These backups will be stored here. Best practices ============== This section describes the recommended approach to configuring these locations, and calls out particular problematic configurations which must be avoided. * **DO** store **Database** on high-throughput, low latency storage, such as an SSD, if available. * **DO** store **Media** on larger, slower storage (e.g., HDDs or network storage), if available. * **DO** store **Backups** separately to **Database**. * **DO** take your own backups, ensuring backups are taken for all storage involved and ensuring Camelot is stopped before these backups are taken. * **DO** store your backups separately to all Camelot storage locations. * **DO NOT** store **Database** on storage crossing a WAN (e.g., Google Drive, OneDrive, Dropbox and other online storage). This will severely degrade Camelot's performance and poses a **high risk of unrecoverable data corruption**. * **DO NOT** store **Database** on local network storage unless disk write caching for that volume is disabled. (This is typically not the default.) * **DO NOT** rely on Camelot's backups alone; they are intended for recovery from a very limited set of problems only. * **DO NOT** use "incremental" backups. Any backups taken of **Database** MUST be full backups. * **DO NOT** store dataset data within the folders configured for other datasets. Switching datasets ~~~~~~~~~~~~~~~~~~ Once there are multiple datasets configured, you can move between them using the dataset switcher in the top right hand corner. .. figure:: screenshot/dataset-switcher.png :alt: