Your Usenet provider and indexers

Two things to enter, once each. Everything else configures itself from them.

The short version

**Connections → Usenet provider.** Your provider's server address, your username and

password, and how many connections your account allows.

**Connections → Search sources.** Each indexer's address and its API key.

That is the whole of it. You will not be asked for either again.

Why once is enough

Three engines use your Usenet account: **InfiniDysk** and **AIOStreams** stream from it,

and **NZBGet** downloads from it. Older setups made you type the same credentials into

all three, and then made you work out how to divide your connections between them.

GeekStream takes the entry once and configures all three, giving each its own share of

the connection limit you entered. If the shares added up to more than your account

allows, your provider would start refusing connections — and it refuses them to whichever

engine asks next, which is usually the one trying to play something.

What to enter

### Usenet provider

| Field | What it is |

|---|---|

| Name | Anything you like; it is only a label |

| Server | The address on your provider's account page, such as `news.example.com` |

| Port | Usually 563 for an encrypted connection |

| Use encryption | Leave on unless your provider says otherwise |

| Username and password | The ones for the news server, which are sometimes not your website sign-in |

| Connection limit | **The number your account actually allows** |

| Retention in days | Optional; how far back your provider keeps files |

| Role | Primary, or backup for a second provider |

**The connection limit matters more than it looks.** Enter what your account allows, not

what you think you need. Too high and your provider refuses connections in the middle of

playback; too low and everything is slower than you paid for. It is on your provider's

account page.

**Retention** is worth filling in. GeekStream stops asking for files older than your

provider keeps, so searches do not spend your indexer requests on things nobody could

download. It is passed to the streaming engine and to NZBGet together.

### Search sources

Each indexer needs its address and its API key, both from your account page there.

You are also asked how many requests that account allows per hour and per day. Enter what

your indexer documents. GeekStream stays inside it: those limits exist to stop an account

being suspended, and a suspended indexer helps nobody. See

[Will my indexer account get rate-limited or suspended?](indexer-account-limits.md).

You can save several. They are tried in the priority order you give them.

After saving

The **Server** tile on Overview stops saying *Waiting for your Usenet provider*.

Then, on the server itself:

sudo bash /opt/geekstream/installer/linux/install.sh --activate-route

This registers the route your television plays through. It is separate from the install

because it plays a real file to prove the route works, which needs a provider and an

indexer saved first.

If a save does not work

**"The storage engine is not ready yet, so nothing was saved."** One of the engines was

still starting. Wait a moment and save again; your existing settings are untouched.

**A screen that will not respond.** Every screen has its own **Check again** button,

which tells you what is wrong with that screen rather than sending you elsewhere.

Nothing you enter here is ever shown back to you, written to a log, or included in a

troubleshooting report. See [Privacy and security](privacy-security.md).