1. Deployment Options

Cerebro uses modular architecture, i. e. consists of several almost independent components (modules), such as:

  • The database – is a core component. It handles user accounts, projects, task structures, messages, permissions, etc. PostgreSQL (http://www.postgresql.org) serves as the DBMS.
  • The file storage – is a server with Cargador service running on it. Cargador is a utility for Cerebro file management.
  • The mailer – sends Cerebro system notifications via email and processes incoming email messages from Cerebro users. The Cargador service is used for those tasks too, that is why usually the mailer is deployed on the same physical server as the file storage.
  • The web server – a server with special software running on it, which allows to use Cerebro via a web browser, without a desktop client application.
  • The client application – is a desktop software that runs on users’ workstations and interacts with server components (database and file storage).

The available Cerebro deployment options may vary from a completely cloud-based service to a completely local (on-premises) client-server application.

There are several basic deployment configurations, depending on a selected pricing plan (https://cerebrohq.com/en/pricing/) – they are described below.

1.1. “Cloud-based” service

Offered by the Home pricing plan.

Using this configuration means that all the server components (database, file storage, mailer and web server) are running on the cerebrohq.com servers and accessible via Internet.

Benefits.

  • Basically, the whole Cerebro installation procedure is skipped (see below).
  • No system administration needed.
  • At the same time, the system is always up-to-date and fully functional, due to the centralized server components.

Restrictions.

  • Stable Internet connection required to uphold database and file storage transactions.
  • The total amount of user accounts is limited to 5.
  • The file storage capacity available is limited to 15 GB.

In this case Cerebro deployment implies only settiung up client applications on users’ workstations (see chapter: “Installing Cerebro On-Premises”).

1.2. Hybrid Cloud Configuration

Offered by Studio pricing plan, as well as by free trial version.

This configuration implies that some of the server components (database, mailer and web server) are running on the cerebrohq.com servers, while the file storage is deployed on an application server within your company’s LAN.

The default component layout:

_images/SaaS_Studio_Plan.png

Benefits.

  • Minimal deployment time.
  • Higher level of stored data security (due to the local file storage, no files are transferred to Cerebro service provider, always remaining under control of the client company).
  • At the same time, the system always remains up-to-date and fully functional, the maintenace costs and effort are minimal (due to the centralized server components).

Restrictions.

  • Constant Internet connection required to uphold transactions with the database.

In this case the Cerebro deployment consists of the following steps:

1.3. Local (On-premises) Deployment

Offered by Factory and Corporation pricing plans.

This configuration implies that all of the server components (database, file storage, mailer and web server) are running and mainained by client company either on-premises or in co-location.

Note

The client company is free to opt for the number and location of the server components. For example, it may deploy more than one Cargador instance, or it may exclude the web server from the configuration.

The most typical configuration is the following: database, file server and mailer are installed onto one physical machine (the so called Application Server) and a web server, if needed, is installed onto another machine (Web Server).

_images/Installed_Plan.png

Benefits.

  • Full control over all system components and contents.
  • Connection to your services stays active even if Internet connection is down.

Restrictions.

  • Larger deployment and maintenance costs (in comparison with hybrid cloud configuration).

In this case the Cerebro deployment consists of the following steps:

The detailed installation guide on each Cerebro component can be found in other chapters of this manual.

1.4. Technical Specifications for Server machine

Below are the production-proven specifications of servers, recommended for reference.

1.4.1. Main Server (Application Server)

  • always up;
  • RAM – 2 GB min, recommended – 4 GB;
  • CPU – dual-core Intel-compatible;
  • OS – Windows Server 32/64 or Linux 32/64 (Ubuntu, Debian);
  • static external IP address;
  • accessible on a UDP port within LAN (default port 45430);
  • TCP ports 45431, 4080 (for Cargador) and 5432 (for Memoria) open and forwarded to Internet;
  • Memoria DB requires PostgreSQL 9.0 (don’t use 9.1 or newer versions).

1.4.2. Web Server

If you plan to use Cerebro web interface, it is recommended to allocate a dedicated server to run it.

  • always up;
  • RAM – 1 GB minimum;
  • CPU – dual-core Intel-compatible;
  • OS – Linux 32/64 (Ubuntu, Debian);
  • static external IP address.