Node working directory
During operation, the system uses the local disk to store service information.
All such resources are placed in the /var directory.
Each node has its own working directory: /var/lib/era/_workdir/<NODE>.
Each node has access to local directories subject to various types of synchronization: /var/lib/era_files.
Each such directory is addressed uniformly in scripts (see Local paths) and is used in various processes by the system’s internal services.
Each node maintains logs: /var/log/era/<NODE>.
If the system runs in a Docker container, the specified paths are available only inside it. The working directory is located on an external volume and remains on disk after the Docker container is removed.
Directory contents
| Resource | Purpose |
|---|---|
|
Node working directory. Each node is named automatically based on the names of the logical roles running on it and the server IP address. The node working directories contain the resources of the running roles. For example, in the diagram, the mdc role stores the license file ( The working directory also contains a symlink to the directory with logs and traces. |
|
Local directories for various types of synchronization. Each such directory is addressed uniformly in scripts (see Local paths) and is used in various processes by the system’s internal services.
When creating a configuration, the paths to these service directory categories can be changed individually on each server. The location given here is recommended and is used by default. |
|
Directory with the node’s logs. It also contains virtual machine process traces when used for debugging: |
|
Directory containing pipes and sockets of the active node. For debugging purposes, pipes are used to connect to the node in the Erlang virtual machine interpreter. |