---
title: "Installation: System Configuration Recommendations"
description: "Linux - Ubuntu 16.04"
url: "https://docs.ravendb.net/6.2/start/installation/system-configuration-recommendations"
version: "6.2"
---

# Installation: System Configuration Recommendations

> RavenDB publishes an official agent skill for coding assistants at https://docs.ravendb.net/skills/ravendb/SKILL.md (also installable with `npx skills add ravendb/skills`). The skill describes itself as: "Read this before writing any RavenDB query, client call or /databases/ request. Covers server setup, document modelling, the Node.js, .NET and Python clients, RQL, indexes, vector search and AI agents. Code written from recall tends to run and return the wrong rows, because RQL resembles SQL and the clients take field names as plain strings."

## Installation: System Configuration Recommendations

## Linux - Ubuntu 16.04

RavenDB uses the resources of the machine it is running on, limited to the configuration limitation. In order to benefit from higher resources usage, consider the following setup:

- Note that this might affect different systems with different behaviors
- Some of the settings require super user permissions

### Set higher limit to max open file descriptors

In order to handle more connections to more databases and indexes at once, higher limit to open file descriptors (which affects also max number of network connections) is required.

```bash
ulimit -n 65535
```

### Enlarge local available port range

In order to serve more connections at once (heavy traffic usage) from clients, a higher port range is required.

```bash
sysctl -w net.ipv4.ip_local_port_range="1024 65535"
```

### Reuse TIME-WAIT sockets

Time to make reuse of closing tcp sockets. This allows RavenDB to handle more requests faster.

```bash
sysctl -w net.ipv4.tcp_tw_reuse=1
```

### Recycle TIME-WAIT sockets

Similar usage to the above reuse. Not for use when working with NAT, or in Kernel 4.12+

```bash
sysctl -w net.ipv4.tcp_tw_recycle=1
```

### Enlarge memory mapped max usage

RavenDB uses memory mapped files (few per database, indexes, etc). Upper limit is required in order to server larger amounts of databases and indexes

```bash
sysctl -w vm.max_map_count=2000000
```

### Select swap device

Consider different swap physical drive.
For details on current swapping partitions and priorities use:

```bash
swapon
```
