Tableau Server Client File Service
The Client File Service (CFS) stores and distributes files needed by TSM (e.g. certificates, customization files, etc.). Files that are managed by the Client File Service are renamed and obfuscated before they are distributed across the deployment. This process also parametrizes the file attributes that are required by Tableau services. As a result, files are not mapped to a single file location on the file system. Be sure you have an off-box backup of all files managed by CFS.
The following files are managed by CFS:
- SAML certificate file
- SAML key file
- SAML IdP metadata file
- The custom certificate installed by tsm security custom-cert add
- Kerberos.keytab file
- LDAP Kerberos keytab file
- LDAP Kerberos conf file
- Mutual SSL certificate file
- Mutual SSL revocation file
- Customization header logo file
- Customization sign-in logo file
- Customization compact logo file
The following files are not managed or distributed by CFS:
- External SSL files. The certificate and key files for external SSL are stored and managed by the Coordination Service. You do not need to manually distribute these files.
- SSL files for LDAP external identity store. You must distribute the SSL certificate file manually to each node in the cluster. See Configure Encrypted Channel to LDAP External Identity Store.
The Client File Service functions much like the File Store does for files needed by business services. By default, CFS is only installed on the initial node of your Tableau Server installation. To configure CFS to for high availability, we recommend that you configure an instance of CFS on each of the nodes where you deploy the Coordination Service.
In a cluster, if a node that is running your only instance of CFS fails, any files being managed by CFS will be lost, and you will need to repopulate CFS those files by reimporting certs and custom images, and making any related configuration changes.
|Status||Status of the
|Logging||Logs generated by the
What happens when a CFS process fails? Nothing, as long as there is still at least one functioning CFS process in the cluster. The controller will redirect file transfer requests to the other working CFS process.
Failed Client File Service processes automatically restart as long as the computer itself is otherwise healthy.