diff options
Diffstat (limited to 'docs/netdata-agent/configuration')
16 files changed, 2493 insertions, 0 deletions
diff --git a/docs/netdata-agent/configuration/README.md b/docs/netdata-agent/configuration/README.md new file mode 100644 index 000000000..097fb9310 --- /dev/null +++ b/docs/netdata-agent/configuration/README.md @@ -0,0 +1,43 @@ +# Netdata Agent Configuration + +The main Netdata agent configuration is `netdata.conf`. + +## The Netdata config directory + +On most Linux systems, by using our [recommended one-line installation](/packaging/installer/README.md#install-on-linux-with-one-line-installer), the **Netdata config +directory** will be `/etc/netdata/`. The config directory contains several configuration files with the `.conf` extension, a +few directories, and a shell script named `edit-config`. + +> Some operating systems will use `/opt/netdata/etc/netdata/` as the config directory. If you're not sure where yours +> is, navigate to `http://NODE:19999/netdata.conf` in your browser, replacing `NODE` with the IP address or hostname of +> your node, and find the `# config directory = ` setting. The value listed is the config directory for your system. + +All of Netdata's documentation assumes that your config directory is at `/etc/netdata`, and that you're running any scripts from inside that directory. + + +## edit `netdata.conf` + +To edit `netdata.conf`, run this on your terminal: + +```bash +cd /etc/netdata 2>/dev/null || cd /opt/netdata/etc/netdata +sudo ./edit-config netdata.conf +``` + +Your editor will open. + +## downloading `netdata.conf` + +The running version of `netdata.conf` can be downloaded from a running Netdata agent, at this URL: + +``` +http://agent-ip:19999/netdata.conf +``` + +You can save and use this version, using these commands: + +```bash +cd /etc/netdata 2>/dev/null || cd /opt/netdata/etc/netdata +curl -ksSLo /tmp/netdata.conf.new http://localhost:19999/netdata.conf && sudo mv -i /tmp/netdata.conf.new netdata.conf +``` + diff --git a/docs/netdata-agent/configuration/anonymous-telemetry-events.md b/docs/netdata-agent/configuration/anonymous-telemetry-events.md new file mode 100644 index 000000000..b943ea9a3 --- /dev/null +++ b/docs/netdata-agent/configuration/anonymous-telemetry-events.md @@ -0,0 +1,103 @@ +<!-- +title: "Anonymous telemetry events" +custom_edit_url: https://github.com/netdata/netdata/edit/master/docs/netdata-agent/configuration/anonymous-telemetry-events.md +sidebar_label: "Anonymous telemetry events" +learn_status: "Published" +learn_rel_path: "Configuration" +--> + +# Anonymous telemetry events + +By default, Netdata collects anonymous usage information from the open-source monitoring agent. For agent events like start,stop,crash etc we use our own cloud function in GCP. For frontend telemetry (pageviews etc.) on the agent dashboard itself we use the open-source +product analytics platform [PostHog](https://github.com/PostHog/posthog). + +We are strongly committed to your [data privacy](https://netdata.cloud/privacy/). + +We use the statistics gathered from this information for two purposes: + +1. **Quality assurance**, to help us understand if Netdata behaves as expected, and to help us classify repeated + issues with certain distributions or environments. + +2. **Usage statistics**, to help us interpret how people use the Netdata agent in real-world environments, and to help + us identify how our development/design decisions influence the community. + +Netdata collects usage information via two different channels: + +- **Agent dashboard**: We use the [PostHog JavaScript integration](https://posthog.com/docs/integrations/js-integration) (with sensitive event attributes overwritten to be anonymized) to send product usage events when you access an [Agent's dashboard](/docs/dashboards-and-charts/README.md). +- **Agent backend**: The `netdata` daemon executes the [`anonymous-statistics.sh`](https://github.com/netdata/netdata/blob/6469cf92724644f5facf343e4bdd76ac0551a418/daemon/anonymous-statistics.sh.in) script when Netdata starts, stops cleanly, or fails. + +You can opt-out from sending anonymous statistics to Netdata through three different [opt-out mechanisms](#opt-out). + +## Agent Dashboard - PostHog JavaScript + +When you kick off an Agent dashboard session by visiting `http://NODE:19999`, Netdata initializes a PostHog session and masks various event attributes. + +_Note_: You can see the relevant code in the [dashboard repository](https://github.com/netdata/dashboard/blob/master/src/domains/global/sagas.ts#L107) where the `window.posthog.register()` call is made. + +```JavaScript +window.posthog.register({ + distinct_id: machineGuid, + $ip: "127.0.0.1", + $current_url: "agent dashboard", + $pathname: "netdata-dashboard", + $host: "dashboard.netdata.io", +}) +``` + +In the above snippet a Netdata PostHog session is initialized and the `ip`, `current_url`, `pathname` and `host` attributes are set to constant values for all events that may be sent during the session. This way, information like the IP or hostname of the Agent will not be sent as part of the product usage event data. + +We have configured the dashboard to trigger the PostHog JavaScript code only when the variable `anonymous_statistics` is true. The value of this +variable is controlled via the [opt-out mechanism](#opt-out). + +## Agent Backend - Anonymous Statistics Script + +Every time the daemon is started or stopped and every time a fatal condition is encountered, Netdata uses the anonymous +statistics script to collect system information and send it to the Netdata telemetry cloud function via an http call. The information collected for all +events is: + +- Netdata version +- OS name, version, id, id_like +- Kernel name, version, architecture +- Virtualization technology +- Containerization technology + +Furthermore, the FATAL event sends the Netdata process & thread name, along with the source code function, source code +filename and source code line number of the fatal error. + +Starting with v1.21, we additionally collect information about: + +- Failures to build the dependencies required to use Cloud features. +- Unavailability of Cloud features in an agent. +- Failures to connect to the Cloud in case the [connection process](/src/claim/README.md) has been completed. This includes error codes + to inform the Netdata team about the reason why the connection failed. + +To see exactly what and how is collected, you can review the script template `daemon/anonymous-statistics.sh.in`. The +template is converted to a bash script called `anonymous-statistics.sh`, installed under the Netdata `plugins +directory`, which is usually `/usr/libexec/netdata/plugins.d`. + +## Opt-out + +You can opt-out from sending anonymous statistics to Netdata through three different opt-out mechanisms: + +**Create a file called `.opt-out-from-anonymous-statistics`.** This empty file, stored in your Netdata configuration +directory (usually `/etc/netdata`), immediately stops the statistics script from running, and works with any type of +installation, including manual, offline, and macOS installations. Create the file by running `touch +.opt-out-from-anonymous-statistics` from your Netdata configuration directory. + +**Pass the option `--disable-telemetry` to any of the installer scripts in the [installation +docs](/packaging/installer/README.md).** You can append this option during the initial installation or a manual +update. You can also export the environment variable `DISABLE_TELEMETRY` with a non-zero or non-empty value +(e.g: `export DISABLE_TELEMETRY=1`). + +When using Docker, **set your `DISABLE_TELEMETRY` environment variable to `1`.** You can set this variable with the following +command: `export DISABLE_TELEMETRY=1`. When creating a container using Netdata's [Docker +image](/packaging/docker/README.md#create-a-new-netdata-agent-container) for the first time, this variable will disable +the anonymous statistics script inside of the container. + +Each of these opt-out processes does the following: + +- Prevents the daemon from executing the anonymous statistics script. +- Forces the anonymous statistics script to exit immediately. +- Stops the PostHog JavaScript snippet, which remains on the dashboard, from firing and sending any data to the Netdata PostHog. + + diff --git a/docs/netdata-agent/configuration/cheatsheet.md b/docs/netdata-agent/configuration/cheatsheet.md new file mode 100644 index 000000000..3e1428694 --- /dev/null +++ b/docs/netdata-agent/configuration/cheatsheet.md @@ -0,0 +1,215 @@ +# Useful management and configuration actions + +Below you will find some of the most common actions that one can take while using Netdata. You can use this page as a quick reference for installing Netdata, connecting a node to the Cloud, properly editing the configuration, accessing Netdata's API, and more! + +### Install Netdata + +```bash +wget -O /tmp/netdata-kickstart.sh https://get.netdata.cloud/kickstart.sh && sh /tmp/netdata-kickstart.sh + +# Or, if you have cURL but not wget (such as on macOS): +curl https://get.netdata.cloud/kickstart.sh > /tmp/netdata-kickstart.sh && sh /tmp/netdata-kickstart.sh +``` + +#### Connect a node to Netdata Cloud + +To do so, sign in to Netdata Cloud, on your Space under the Nodes tab, click `Add Nodes` and paste the provided command into your node’s terminal and run it. +You can also copy the Claim token and pass it to the installation script with `--claim-token` and re-run it. + +### Configuration + +**Netdata's config directory** is `/etc/netdata/` but in some operating systems it might be `/opt/netdata/etc/netdata/`. +Look for the `# config directory =` line over at `http://NODE_IP:19999/netdata.conf` to find your config directory. + +From within that directory you can run `sudo ./edit-config netdata.conf` **to edit Netdata's configuration.** +You can edit other config files too, by specifying their filename after `./edit-config`. +You are expected to use this method in all following configuration changes. + +<!-- #### Edit Netdata's other config files (examples): + +- `$ sudo ./edit-config apps_groups.conf` +- `$ sudo ./edit-config ebpf.conf` +- `$ sudo ./edit-config health.d/load.conf` +- `$ sudo ./edit-config go.d/prometheus.conf` + +#### View the running Netdata configuration: `http://NODE:19999/netdata.conf` + +> Replace `NODE` with the IP address or hostname of your node. Often `localhost`. + +## Metrics collection & retention + +You can tweak your settings in the netdata.conf file. +📄 [Find your netdata.conf file](/src/daemon/config/README.md) + +Open a new terminal and navigate to the netdata.conf file. Use the edit-config script to make changes: `sudo ./edit-config netdata.conf` + +The most popular settings to change are: + +#### Increase metrics retention (4GiB) + +``` +sudo ./edit-config netdata.conf +``` + +``` +[global] + dbengine multihost disk space = 4096 +``` + +#### Reduce the collection frequency (every 5 seconds) + +``` +sudo ./edit-config netdata.conf +``` + +``` +[global] + update every = 5 +``` --> + +--- + +#### Enable/disable plugins (groups of collectors) + +```bash +sudo ./edit-config netdata.conf +``` + +```conf +[plugins] + go.d = yes # enabled + node.d = no # disabled +``` + +#### Enable/disable specific collectors + +```bash +sudo ./edit-config go.d.conf # edit a plugin's config +``` + +```yaml +modules: + activemq: no # disabled + cockroachdb: yes # enabled +``` + +#### Edit a collector's config + +```bash +sudo ./edit-config go.d/mysql.conf +``` + +### Alerts & notifications + +<!-- #### Add a new alert + +``` +sudo touch health.d/example-alert.conf +sudo ./edit-config health.d/example-alert.conf +``` --> +After any change, reload the Netdata health configuration: + +```bash +netdatacli reload-health +#or if that command doesn't work on your installation, use: +killall -USR2 netdata +``` + +#### Configure a specific alert + +```bash +sudo ./edit-config health.d/example-alert.conf +``` + +#### Silence a specific alert + +```bash +sudo ./edit-config health.d/example-alert.conf +``` + +``` + to: silent +``` + +<!-- #### Disable alerts and notifications + +```conf +[health] + enabled = no +``` --> + +--- + +### Manage the daemon + +| Intent | Action | +|:----------------------------|------------------------------------------------------------:| +| Start Netdata | `$ sudo service netdata start` | +| Stop Netdata | `$ sudo service netdata stop` | +| Restart Netdata | `$ sudo service netdata restart` | +| Reload health configuration | `$ sudo netdatacli reload-health` `$ killall -USR2 netdata` | +| View error logs | `less /var/log/netdata/error.log` | +| View collectors logs | `less /var/log/netdata/collector.log` | + +#### Change the port Netdata listens to (example, set it to port 39999) + +```conf +[web] +default port = 39999 +``` + +### See metrics and dashboards + +#### Netdata Cloud: `https://app.netdata.cloud` + +#### Local dashboard: `https://NODE:19999` + +> Replace `NODE` with the IP address or hostname of your node. Often `localhost`. + +### Access the Netdata API + +You can access the API like this: `http://NODE:19999/api/VERSION/REQUEST`. +If you want to take a look at all the API requests, check our API page at <https://learn.netdata.cloud/api> +<!-- +## Interact with charts + +| Intent | Action | +| -------------------------------------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: | +| Stop a chart from updating | `click` | +| Zoom | **Cloud** <br/> use the `zoom in` and `zoom out` buttons on any chart (upper right corner) <br/><br/> **Agent**<br/>`SHIFT` or `ALT` + `mouse scrollwheel` <br/> `SHIFT` or `ALT` + `two-finger pinch` (touchscreen) <br/> `SHIFT` or `ALT` + `two-finger scroll` (touchscreen) | +| Zoom to a specific timeframe | **Cloud**<br/>use the `select and zoom` button on any chart and then do a `mouse selection` <br/><br/> **Agent**<br/>`SHIFT` + `mouse selection` | +| Pan forward or back in time | `click` & `drag` <br/> `touch` & `drag` (touchpad/touchscreen) | +| Select a certain timeframe | `ALT` + `mouse selection` <br/> WIP need to evaluate this `command?` + `mouse selection` (macOS) | +| Reset to default auto refreshing state | `double click` | --> + +<!-- ## Dashboards + +#### Disable the local dashboard + +Use the `edit-config` script to edit the `netdata.conf` file. + +``` +[web] +mode = none +``` --> + +<!-- #### Opt out from anonymous statistics + +``` +sudo touch .opt-out-from-anonymous-statistics +``` --> + +<!-- ## Understanding the dashboard + +**Charts**: A visualization displaying one or more collected/calculated metrics in a time series. Charts are generated +by collectors. + +**Dimensions**: Any value shown on a chart, which can be raw or calculated values, such as percentages, averages, +minimums, maximums, and more. + +**Families**: One instance of a monitored hardware or software resource that needs to be monitored and displayed +separately from similar instances. Example, disks named +**sda**, **sdb**, **sdc**, and so on. + +**Contexts**: A grouping of charts based on the types of metrics collected and visualized. +**disk.io**, **disk.ops**, and **disk.backlog** are all contexts. --> diff --git a/docs/netdata-agent/configuration/common-configuration-changes.md b/docs/netdata-agent/configuration/common-configuration-changes.md new file mode 100644 index 000000000..e9d8abadc --- /dev/null +++ b/docs/netdata-agent/configuration/common-configuration-changes.md @@ -0,0 +1,148 @@ +# Common configuration changes + +The Netdata Agent requires no configuration upon installation to collect thousands of per-second metrics from most +systems, containers, and applications, but there are hundreds of settings to tweak if you want to exercise more control +over your monitoring platform. + +This document assumes familiarity with +using [`edit-config`](/docs/netdata-agent/configuration/README.md) from the Netdata config +directory. + +## Change dashboards and visualizations + +The Netdata Agent's [local dashboard](/docs/dashboards-and-charts/README.md), accessible +at `http://NODE:19999` is highly configurable. If +you use [Netdata Cloud](/docs/netdata-cloud/README.md) +for infrastructure monitoring, you +will see many of these +changes reflected in those visualizations due to the way Netdata Cloud proxies metric data and metadata to your browser. + +### Increase the long-term metrics retention period + +Read our doc +on [increasing long-term metrics storage](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md) +for details, including a +[calculator](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md#calculate-the-system-resources-ram-disk-space-needed-to-store-metrics) +to help you determine the exact settings for your desired retention period. + +### Reduce the data collection frequency + +Change `update every` in +the [`[global]` section](/src/daemon/config/README.md#global-section-options) +of `netdata.conf` so +that it is greater than `1`. An `update every` of `5` means the Netdata Agent enforces a _minimum_ collection frequency +of 5 seconds. + +```conf +[global] + update every = 5 +``` + +Every collector and plugin has its own `update every` setting, which you can also change in the `go.d.conf`, +`python.d.conf` or `charts.d.conf` files, or in individual collector configuration files. If the `update +every` for an individual collector is less than the global, the Netdata Agent uses the global setting. See +the [enable or configure a collector](/src/collectors/REFERENCE.md#enable-and-disable-a-specific-collection-module) +doc for details. + +### Disable a collector or plugin + +Turn off entire plugins in +the [`[plugins]` section](/src/daemon/config/README.md#plugins-section-options) +of +`netdata.conf`. + +To disable specific collectors, open `go.d.conf`, `python.d.conf` or `charts.d.conf` and find the line +for that specific module. Uncomment the line and change its value to `no`. + +## Modify alerts and notifications + +Netdata's health monitoring watchdog uses hundreds of preconfigured health entities, with intelligent thresholds, to +generate warning and critical alerts for most production systems and their applications without configuration. However, +each alert and notification method is completely customizable. + +### Add a new alert + +To create a new alert configuration file, initiate an empty file, with a filename that ends in `.conf`, in the +`health.d/` directory. The Netdata Agent loads any valid alert configuration file ending in `.conf` in that directory. +Next, edit the new file with `edit-config`. For example, with a file called `example-alert.conf`. + +```bash +sudo touch health.d/example-alert.conf +sudo ./edit-config health.d/example-alert.conf +``` + +Or, append your new alert to an existing file by editing a relevant existing file in the `health.d/` directory. + +Read more about [configuring alerts](/src/health/REFERENCE.md) to +get started, and see +the [health monitoring reference](/src/health/REFERENCE.md) for a full listing +of options available in health entities. + +### Configure a specific alert + +Tweak existing alerts by editing files in the `health.d/` directory. For example, edit `health.d/cpu.conf` to change how +the Agent responds to anomalies related to CPU utilization. + +To see which configuration file you need to edit to configure a specific +alert, [view your active alerts](/docs/dashboards-and-charts/alerts-tab.md) in +Netdata Cloud or the local Agent dashboard and look for the **source** line. For example, it might +read `source 4@/usr/lib/netdata/conf.d/health.d/cpu.conf`. + +Because the source path contains `health.d/cpu.conf`, run `sudo edit-config health.d/cpu.conf` to configure that alert. + +### Disable a specific alert + +Open the configuration file for that alert and set the `to` line to `silent`. + +```conf +template: disk_fill_rate + on: disk.space + lookup: max -1s at -30m unaligned of avail + calc: ($this - $avail) / (30 * 60) + every: 15s + to: silent +``` + +### Turn of all alerts and notifications + +Set `enabled` to `no` in +the [`[health]`](/src/daemon/config/README.md#health-section-options) +section of `netdata.conf`. + +### Enable alert notifications + +Open `health_alarm_notify.conf` for editing. First, read the [enabling notifications](/docs/alerts-and-notifications/notifications/README.md#netdata-agent) doc +for an example of the process using Slack, then +click on the link to your preferred notification method to find documentation for that specific endpoint. + +## Improve node security + +While the Netdata Agent is both [open and secure by design](https://www.netdata.cloud/blog/netdata-agent-dashboard/), we +recommend every user take some action to administer and secure their nodes. + +Learn more about the available options in the [security design documentation](/docs/security-and-privacy-design/README.md). + +## Reduce resource usage + +Read +our [performance optimization guide](/docs/netdata-agent/configuration/optimize-the-netdata-agents-performance.md) +for a long list of specific changes +that can reduce the Netdata Agent's CPU/memory footprint and IO requirements. + +## Organize nodes with host labels + +Beginning with v1.20, Netdata accepts user-defined **host labels**. These labels are sent during streaming, exporting, +and as metadata to Netdata Cloud, and help you organize the metrics coming from complex infrastructure. Host labels are +defined in the section `[host labels]`. + +For a quick introduction, read +the [host label guide](/docs/netdata-agent/configuration/organize-systems-metrics-and-alerts.md). + +The following restrictions apply to host label names: + +- Names cannot start with `_`, but it can be present in other parts of the name. +- Names only accept alphabet letters, numbers, dots, and dashes. + +The policy for values is more flexible, but you can not use exclamation marks (`!`), whitespaces (` `), single quotes +(`'`), double quotes (`"`), or asterisks (`*`), because they are used to compare label values in health alerts and +templates. diff --git a/docs/netdata-agent/configuration/dynamic-configuration.md b/docs/netdata-agent/configuration/dynamic-configuration.md new file mode 100644 index 000000000..7064abf9a --- /dev/null +++ b/docs/netdata-agent/configuration/dynamic-configuration.md @@ -0,0 +1,62 @@ +# Dynamic Configuration Manager + +**Netdata Cloud paid subscription required.** + +The Dynamic Configuration Manager allows direct configuration of collectors and alerts through the Netdata UI. This feature allows users to: + +- **Create, test, and deploy configurations** for one or more nodes directly within the UI. +- **Eliminate the need for manual command-line edits and node access**, enhancing workflow efficiency. + +**Cloud Connection and Security**: Nodes using Dynamic Configuration Manager require a connection to Netdata Cloud. This ensures proper permission handling and data security. + +> **Info** +> +> To understand what actions users can perform based on their role, refer to the [Role Based Access documentation](/docs/netdata-cloud/authentication-and-authorization/role-based-access-model.md#dynamic-configuration-manager). + +## Collectors + +### Module + +A module represents a specific data collector, such as Apache, MySQL, or Redis. Think of modules as templates for data collection. + +Each module can have multiple jobs, which are unique configurations of that template tailored to your specific needs. + +You can manage individual modules using the following actions: + +| Action | Description | +|--------------------|---------------------------------------------------------------------------------------------------------------------------| +| **Add job** | Create new configuration instances (jobs) for a particular module. | +| **Enable/Disable** | Disabling a module deactivates all currently running jobs and prevents any future jobs from being created for that module | + +### Job + +A job represents a running instance of a module with a specific configuration. Think of it as a customized data collection task based on a module template. + +Every job has a designated "source type" indicating its origin: + +- **Stock**: Pre-installed with Netdata and provides basic data collection for common services. +- **User**: Originates from user-created files on the node. +- **Discovered**: Automatically generated by Netdata upon discovering a service running on the node. +- **Dynamic Configuration**: Created and managed using the Dynamic Configuration Manager. + +You can manage individual jobs using the following actions: + +| Action | Description | +|--------------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| +| **Restart** | This restarts a job's data collection, useful if a job encounters a "Failed" state. Upon restart, a notification with the failure message will be displayed. | +| **Remove** | Delete a job configuration entirely. Note that only jobs created through Dynamic Configuration can be removed. Other job types originate from files on the node and cannot be deleted here. | +| **Enable/Disable** | Control the job's activity. Disabling a running job stops data collection. | +| **Edit** | Modify an existing job's configuration. | +| **Test** | Validate newly created or edited configurations before applying them permanently. | + +## Health + +Each entry in the Health tab contains an Alert template, that then is used to create Alerts. + +The functionality in the main view is the same as with the [Collectors tab](#collectors). + +### Health entry configuration + +You can create new configurations both for templates or individual Alerts. + +Each template can have multiple items which resemble Alerts that either apply to a certain [instance](/docs/dashboards-and-charts/netdata-charts.md#instances-dropdown), or all instances under a specific [context](/docs/dashboards-and-charts/netdata-charts.md#contexts) diff --git a/docs/netdata-agent/configuration/optimize-the-netdata-agents-performance.md b/docs/netdata-agent/configuration/optimize-the-netdata-agents-performance.md new file mode 100644 index 000000000..6acbd4977 --- /dev/null +++ b/docs/netdata-agent/configuration/optimize-the-netdata-agents-performance.md @@ -0,0 +1,266 @@ +# How to optimize the Netdata Agent's performance + +We designed the Netdata Agent to be incredibly lightweight, even when it's collecting a few thousand dimensions every +second and visualizing that data into hundreds of charts. However, the default settings of the Netdata Agent are not +optimized for performance, but for a simple, standalone setup. We want the first install to give you something you can +run without any configuration. Most of the settings and options are enabled, since we want you to experience the full +thing. + +By default, Netdata will automatically detect applications running on the node it is installed to start collecting +metrics in real-time, has health monitoring enabled to evaluate alerts and trains Machine Learning (ML) models for each +metric, to detect anomalies. + +This document describes the resources required for the various default capabilities and the strategies to optimize +Netdata for production use. + +## Summary of performance optimizations + +The following table summarizes the effect of each optimization on the CPU, RAM and Disk IO utilization in production. + +| Optimization | CPU | RAM | Disk IO | +|-------------------------------------------------------------------------------------------------------------------------------|--------------------|--------------------|--------------------| +| [Use streaming and replication](#use-streaming-and-replication) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | +| [Disable unneeded plugins or collectors](#disable-unneeded-plugins-or-collectors) | :heavy_check_mark: | :heavy_check_mark: | :heavy_check_mark: | +| [Reduce data collection frequency](#reduce-collection-frequency) | :heavy_check_mark: | | :heavy_check_mark: | +| [Change how long Netdata stores metrics](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md) | | :heavy_check_mark: | :heavy_check_mark: | +| [Use a different metric storage database](/src/database/README.md) | | :heavy_check_mark: | :heavy_check_mark: | +| [Disable machine learning](#disable-machine-learning) | :heavy_check_mark: | | | +| [Use a reverse proxy](#run-netdata-behind-a-proxy) | :heavy_check_mark: | | | +| [Disable/lower gzip compression for the agent dashboard](#disablelower-gzip-compression-for-the-dashboard) | :heavy_check_mark: | | | + +## Resources required by a default Netdata installation + +Netdata's performance is primarily affected by **data collection/retention** and **clients accessing data**. + +You can configure almost all aspects of data collection/retention, and certain aspects of clients accessing data. + +### CPU consumption + +Expect about: + +- 1-3% of a single core for the netdata core +- 1-3% of a single core for the various collectors (e.g. go.d.plugin, apps.plugin) +- 5-10% of a single core, when ML training runs + +Your experience may vary depending on the number of metrics collected, the collectors enabled and the specific +environment they run on, i.e. the work they have to do to collect these metrics. + +As a general rule, for modern hardware and VMs, the total CPU consumption of a standalone Netdata installation, +including all its components, should be below 5 - 15% of a single core. For example, on 8 core server it will use only +0.6% - 1.8% of a total CPU capacity, depending on the CPU characteristics. + +The Netdata Agent runs with the lowest +possible [process scheduling policy](/src/daemon/README.md#netdata-process-scheduling-policy), +which is `nice 19`, and uses the `idle` process scheduler. Together, these settings ensure that the Agent only gets CPU +resources when the node has CPU resources to space. If the node reaches 100% CPU utilization, the Agent is stopped first +to ensure your applications get any available resources. + +To reduce CPU usage you can (either one or a combination of the following actions): + +1. [Disable machine learning](#disable-machine-learning), +2. [Use streaming and replication](#use-streaming-and-replication), +3. [Reduce the data collection frequency](#reduce-collection-frequency) +4. [Disable unneeded plugins or collectors](#disable-unneeded-plugins-or-collectors) +5. [Use a reverse proxy](#run-netdata-behind-a-proxy), +6. [Disable/lower gzip compression for the agent dashboard](#disablelower-gzip-compression-for-the-dashboard). + +### Memory consumption + +The memory footprint of Netdata is mainly influenced by the number of metrics concurrently being collected. Expect about +150MB of RAM for a typical 64-bit server collecting about 2000 to 3000 metrics. + +To estimate and control memory consumption, you can (either one or a combination of the following actions): + +1. [Disable unneeded plugins or collectors](#disable-unneeded-plugins-or-collectors) +2. [Change how long Netdata stores metrics](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md) +3. [Use a different metric storage database](/src/database/README.md). + +### Disk footprint and I/O + +By default, Netdata should not use more than 1GB of disk space, most of which is dedicated for storing metric data and +metadata. For typical installations collecting 2000 - 3000 metrics, this storage should provide a few days of +high-resolution retention (per second), about a month of mid-resolution retention (per minute) and more than a year of +low-resolution retention (per hour). + +Netdata spreads I/O operations across time. For typical standalone installations there should be a few write operations +every 5-10 seconds of a few kilobytes each, occasionally up to 1MB. In addition, under heavy load, collectors that +require disk I/O may stop and show gaps in charts. + +To optimize your disk footprint in any aspect described below you can: + + +To configure retention, you can: + +1. [Change how long Netdata stores metrics](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md). + +To control disk I/O: + +1. [Use a different metric storage database](/src/database/README.md), + + +Minimize deployment impact on the production system by optimizing disk footprint: + +1. [Using streaming and replication](#use-streaming-and-replication) +2. [Reduce the data collection frequency](#reduce-collection-frequency) +3. [Disable unneeded plugins or collectors](#disable-unneeded-plugins-or-collectors). + +## Use streaming and replication + +For all production environments, parent Netdata nodes outside the production infrastructure should be receiving all +collected data from children Netdata nodes running on the production infrastructure, +using [streaming and replication](/docs/observability-centralization-points/README.md). + +### Disable health checks on the child nodes + +When you set up streaming, we recommend you run your health checks on the parent. This saves resources on the children +and makes it easier to configure or disable alerts and agent notifications. + +The parents by default run health checks for each child, as long as the child is connected (the details are +in `stream.conf`). On the child nodes you should add to `netdata.conf` the following: + +```conf +[health] + enabled = no +``` + +### Use memory mode ram for the child nodes + +See [using a different metric storage database](/src/database/README.md). + +## Disable unneeded plugins or collectors + +If you know that you don't need an [entire plugin or a specific +collector](/src/collectors/README.md#collector-architecture-and-terminology), +you can disable any of them. Keep in mind that if a plugin/collector has nothing to do, it simply shuts down and does +not consume system resources. You will only improve the Agent's performance by disabling plugins/collectors that are +actively collecting metrics. + +Open `netdata.conf` and scroll down to the `[plugins]` section. To disable any plugin, uncomment it and set the value to +`no`. For example, to explicitly keep the `proc` and `go.d` plugins enabled while disabling `python.d` and `charts.d`. + +```conf +[plugins] + proc = yes + python.d = no + charts.d = no + go.d = yes +``` + +Disable specific collectors by opening their respective plugin configuration files, uncommenting the line for the +collector, and setting its value to `no`. + +```bash +sudo ./edit-config go.d.conf +sudo ./edit-config python.d.conf +sudo ./edit-config charts.d.conf +``` + +For example, to disable a few Python collectors: + +```conf +modules: + apache: no + dockerd: no + fail2ban: no +``` + +## Reduce collection frequency + +The fastest way to improve the Agent's resource utilization is to reduce how often it collects metrics. + +### Global + +If you don't need per-second metrics, or if the Netdata Agent uses a lot of CPU even when no one is viewing that node's +dashboard, [configure the Agent](/docs/netdata-agent/configuration/README.md) to collect +metrics less often. + +Open `netdata.conf` and edit the `update every` setting. The default is `1`, meaning that the Agent collects metrics +every second. + +If you change this to `2`, Netdata enforces a minimum `update every` setting of 2 seconds, and collects metrics every +other second, which will effectively halve CPU utilization. Set this to `5` or `10` to collect metrics every 5 or 10 +seconds, respectively. + +```conf +[global] + update every = 5 +``` + +### Specific plugin or collector + +Every collector and plugin has its own `update every` setting, which you can also change in the `go.d.conf`, +`python.d.conf`, or `charts.d.conf` files, or in individual collector configuration files. If the `update +every` for an individual collector is less than the global, the Netdata Agent uses the global setting. See +the [collectors configuration reference](/src/collectors/REFERENCE.md) for +details. + +To reduce the frequency of +an [internal_plugin/collector](/src/collectors/README.md#collector-architecture-and-terminology), +open `netdata.conf` and find the appropriate section. For example, to reduce the frequency of the `apps` plugin, which +collects and visualizes metrics on application resource utilization: + +```conf +[plugin:apps] + update every = 5 +``` + +To [configure an individual collector](/src/collectors/REFERENCE.md#configure-a-collector), +open its specific configuration file with `edit-config` and look for the `update_every` setting. For example, to reduce +the frequency of the `nginx` collector, run `sudo ./edit-config go.d/nginx.conf`: + +```conf +# [ GLOBAL ] +update_every: 10 +``` + +## Lower memory usage for metrics retention + +See how +to [change how long Netdata stores metrics](/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md). + +## Use a different metric storage database + +Consider [using a different metric storage database](/src/database/README.md) +when running Netdata on IoT devices, and for children in a parent-child set up based +on [streaming and replication](/docs/observability-centralization-points/README.md). + +## Disable machine learning + +Automated anomaly detection may be a powerful tool, but we recommend it to only be enabled on Netdata parents that sit +outside your production infrastructure, or if you have cpu and memory to spare. You can disable ML with the following: + +```conf +[ml] + enabled = no +``` + +## Run Netdata behind a proxy + +A dedicated web server like nginx provides more robustness than the Agent's +internal [web server](/src/web/README.md). +Nginx can handle more concurrent connections, reuse idle connections, and use fast gzip compression to reduce payloads. + +For details on installing another web server as a proxy for the local Agent dashboard, +see [reverse proxies](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/README.md). + +## Disable/lower gzip compression for the dashboard + +If you choose not to run the Agent behind Nginx, you can disable or lower the Agent's web server's gzip compression. +While gzip compression does reduce the size of the HTML/CSS/JS payload, it does use additional CPU while a user is +looking at the local Agent dashboard. + +To disable gzip compression, open `netdata.conf` and find the `[web]` section: + +```conf +[web] + enable gzip compression = no +``` + +Or to lower the default compression level: + +```conf +[web] + enable gzip compression = yes + gzip compression level = 1 +``` + diff --git a/docs/netdata-agent/configuration/optimizing-metrics-database/README.md b/docs/netdata-agent/configuration/optimizing-metrics-database/README.md new file mode 100644 index 000000000..fdbd3b690 --- /dev/null +++ b/docs/netdata-agent/configuration/optimizing-metrics-database/README.md @@ -0,0 +1,3 @@ +# Optimizing Metrics Database Overview + +This section contains documentation to help you understand how the metrics DB works, understand the key features and configure them to suit your needs.
\ No newline at end of file diff --git a/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md b/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md new file mode 100644 index 000000000..8a8659eff --- /dev/null +++ b/docs/netdata-agent/configuration/optimizing-metrics-database/change-metrics-storage.md @@ -0,0 +1,124 @@ +# Change how long Netdata stores metrics + +Netdata offers a granular approach to data retention, allowing you to manage storage based on both **time** and **disk +space**. This provides greater control and helps you optimize storage usage for your specific needs. + +**Default Retention Limits**: + +| Tier | Resolution | Time Limit | Size Limit (min 256 MB) | +|:----:|:-------------------:|:----------:|:-----------------------:| +| 0 | high (per second) | 14 days | 1 GiB | +| 1 | middle (per minute) | 3 months | 1 GiB | +| 2 | low (per hour) | 2 years | 1 GiB | + +> **Note**: If a user sets a disk space size less than 256 MB for a tier, Netdata will automatically adjust it to 256 MB. + +With these defaults, Netdata requires approximately 4 GiB of storage space (including metadata). + +## Retention Settings + +> **In a parent-child setup**, these settings manage the shared storage space utilized by the Netdata parent agent for +> storing metrics collected by both the parent and its child nodes. + +You can fine-tune retention for each tier by setting a time limit or size limit. Setting a limit to 0 disables it, +allowing for no time-based deletion for that tier or using all available space, respectively. This enables various +retention strategies as shown in the table below: + +| Setting | Retention Behavior | +|--------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------| +| Size Limit = 0, Time Limit > 0 | **Time-based only:** data is stored for a specific duration regardless of disk usage. | +| Time Limit = 0, Size Limit > 0 | **Space-based only:** data is stored until it reaches a certain amount of disk space, regardless of time. | +| Time Limit > 0, Size Limit > 0 | **Combined time and space limits:** data is deleted once it reaches either the time limit or the disk space limit, whichever comes first. | + +You can change these limits in `netdata.conf`: + +``` +[db] + mode = dbengine + storage tiers = 3 + + # Tier 0, per second data. Set to 0 for no limit. + dbengine tier 0 disk space MB = 1024 + dbengine tier 0 retention days = 14 + + # Tier 1, per minute data. Set to 0 for no limit. + dbengine tier 1 disk space MB = 1024 + dbengine tier 1 retention days = 90 + + # Tier 2, per hour data. Set to 0 for no limit. + dbengine tier 2 disk space MB = 1024 + dbengine tier 2 retention days = 730 +``` + +## Monitoring Retention Utilization + +Netdata provides a visual representation of storage utilization for both time and space limits across all tiers within +the 'dbengine retention' subsection of the 'Netdata Monitoring' section on the dashboard. This chart shows exactly how +your storage space (disk space limits) and time (time limits) are used for metric retention. + +## Legacy configuration + +### v1.45.6 and prior + +Netdata versions prior to v1.46.0 relied on a disk space-based retention. + +**Default Retention Limits**: + +| Tier | Resolution | Size Limit | +|:----:|:-------------------:|:----------:| +| 0 | high (per second) | 256 MB | +| 1 | middle (per minute) | 128 MB | +| 2 | low (per hour) | 64 GiB | + +You can change these limits in `netdata.conf`: + +``` +[db] + mode = dbengine + storage tiers = 3 + + # Tier 0, per second data + dbengine multihost disk space MB = 256 + + # Tier 1, per minute data + dbengine tier 1 multihost disk space MB = 1024 + + # Tier 2, per hour data + dbengine tier 2 multihost disk space MB = 1024 +``` + +### v1.35.1 and prior + +These versions of the Agent do not support tiers. You could change the metric retention for the parent and +all of its children only with the `dbengine multihost disk space MB` setting. This setting accounts the space allocation +for the parent node and all of its children. + +To configure the database engine, look for the `page cache size MB` and `dbengine multihost disk space MB` settings in +the `[db]` section of your `netdata.conf`. + +```conf +[db] + dbengine page cache size MB = 32 + dbengine multihost disk space MB = 256 +``` + +### v1.23.2 and prior + +_For Netdata Agents earlier than v1.23.2_, the Agent on the parent node uses one dbengine instance for itself, and +another instance for every child node it receives metrics from. If you had four streaming nodes, you would have five +instances in total (`1 parent + 4 child nodes = 5 instances`). + +The Agent allocates resources for each instance separately using the `dbengine disk space MB` (**deprecated**) setting. +If `dbengine disk space MB`(**deprecated**) is set to the default `256`, each instance is given 256 MiB in disk space, +which means the total disk space required to store all instances is, +roughly, `256 MiB * 1 parent * 4 child nodes = 1280 MiB`. + +#### Backward compatibility + +All existing metrics belonging to child nodes are automatically converted to legacy dbengine instances and the localhost +metrics are transferred to the multihost dbengine instance. + +All new child nodes are automatically transferred to the multihost dbengine instance and share its page cache and disk +space. If you want to migrate a child node from its legacy dbengine instance to the multihost dbengine instance, you +must delete the instance's directory, which is located in `/var/cache/netdata/MACHINE_GUID/dbengine`, after stopping the +Agent. diff --git a/docs/netdata-agent/configuration/organize-systems-metrics-and-alerts.md b/docs/netdata-agent/configuration/organize-systems-metrics-and-alerts.md new file mode 100644 index 000000000..b0094a60f --- /dev/null +++ b/docs/netdata-agent/configuration/organize-systems-metrics-and-alerts.md @@ -0,0 +1,253 @@ +# Organize systems, metrics, and alerts + +When you use Netdata to monitor and troubleshoot an entire infrastructure, you need sophisticated ways of keeping everything organized. +Netdata allows to organize your observability infrastructure with Spaces, Rooms, virtual nodes, host labels, and metric labels. + +## Spaces and Rooms + +[Spaces](/docs/netdata-cloud/organize-your-infrastructure-invite-your-team.md#netdata-cloud-spaces) are used for organization-level or infrastructure-level +grouping of nodes and people. A node can only appear in a single space, while people can have access to multiple spaces. + +The [Rooms](/docs/netdata-cloud/organize-your-infrastructure-invite-your-team.md#netdata-cloud-rooms) in a space bring together nodes and people in +collaboration areas. Rooms can also be used for fine-tuned +[role based access control](/docs/netdata-cloud/authentication-and-authorization/role-based-access-model.md). + +## Virtual nodes + +Netdata’s virtual nodes functionality allows you to define nodes in configuration files and have them be treated as regular nodes +in all of the UI, dashboards, tabs, filters etc. For example, you can create a virtual node each for all your Windows machines +and monitor them as discrete entities. Virtual nodes can help you simplify your infrastructure monitoring and focus on the +individual node that matters. + +To define your windows server as a virtual node you need to: + + * Define virtual nodes in `/etc/netdata/vnodes/vnodes.conf` + + ```yaml + - hostname: win_server1 + guid: <value> + ``` + Just remember to use a valid guid (On Linux you can use `uuidgen` command to generate one, on Windows just use the `[guid]::NewGuid()` command in PowerShell) + + * Add the vnode config to the data collection job. e.g. in `go.d/windows.conf`: + ```yaml + jobs: + - name: win_server1 + vnode: win_server1 + url: http://203.0.113.10:9182/metrics + ``` + +## Host labels + +Host labels can be extremely useful when: + +- You need alerts that adapt to the system's purpose +- You need properly-labeled metrics archiving so you can sort, correlate, and mash-up your data to your heart's content. +- You need to keep tabs on ephemeral Docker containers in a Kubernetes cluster. + +Let's take a peek into how to create host labels and apply them across a few of Netdata's features to give you more +organization power over your infrastructure. + +### Default labels + +When Netdata starts, it captures relevant information about the system and converts them into automatically generated +host labels. You can use these to logically organize your systems via health entities, exporting metrics, +parent-child status, and more. + +They capture the following: + +- Kernel version +- Operating system name and version +- CPU architecture, system cores, CPU frequency, RAM, and disk space +- Whether Netdata is running inside of a container, and if so, the OS and hardware details about the container's host +- Whether Netdata is running inside K8s node +- What virtualization layer the system runs on top of, if any +- Whether the system is a streaming parent or child + +If you want to organize your systems without manually creating host labels, try the automatic labels in some of the +features below. You can see them under `http://HOST-IP:19999/api/v1/info`, beginning with an underscore `_`. +```json +{ + ... + "host_labels": { + "_is_k8s_node": "false", + "_is_parent": "false", + ... +``` + +### Custom labels + +Host labels are defined in `netdata.conf`. To create host labels, open that file using `edit-config`. + +```bash +cd /etc/netdata # Replace this path with your Netdata config directory, if different +sudo ./edit-config netdata.conf +``` + +Create a new `[host labels]` section defining a new host label and its value for the system in question. Make sure not +to violate any of the [host label naming rules](/docs/netdata-agent/configuration/common-configuration-changes.md#organize-nodes-with-host-labels). + +```conf +[host labels] + type = webserver + location = us-seattle + installed = 20200218 +``` + +Once you've written a few host labels, you need to enable them. Instead of restarting the entire Netdata service, you +can reload labels using the helpful `netdatacli` tool: + +```bash +netdatacli reload-labels +``` + +Your host labels will now be enabled. You can double-check these by using `curl http://HOST-IP:19999/api/v1/info` to +read the status of your agent. For example, from a VPS system running Debian 10: + +```json +{ + ... + "host_labels": { + "_is_k8s_node": "false", + "_is_parent": "false", + "_virt_detection": "systemd-detect-virt", + "_container_detection": "none", + "_container": "unknown", + "_virtualization": "kvm", + "_architecture": "x86_64", + "_kernel_version": "4.19.0-6-amd64", + "_os_version": "10 (buster)", + "_os_name": "Debian GNU/Linux", + "type": "webserver", + "location": "seattle", + "installed": "20200218" + }, + ... +} +``` + + +### Host labels in streaming + +You may have noticed the `_is_parent` and `_is_child` automatic labels from above. Host labels are also now +streamed from a child to its parent node, which concentrates an entire infrastructure's OS, hardware, container, +and virtualization information in one place: the parent. + +Now, if you'd like to remind yourself of how much RAM a certain child node has, you can access +`http://localhost:19999/host/CHILD_HOSTNAME/api/v1/info` and reference the automatically-generated host labels from the +child system. It's a vastly simplified way of accessing critical information about your infrastructure. + +> ⚠️ Because automatic labels for child nodes are accessible via API calls, and contain sensitive information like +> kernel and operating system versions, you should secure streaming connections with SSL. See the [streaming +> documentation](/src/streaming/README.md#securing-streaming-communications) for details. You may also want to use +> [access lists](/src/web/server/README.md#access-lists) or [expose the API only to LAN/localhost +> connections](/docs/netdata-agent/securing-netdata-agents.md#expose-netdata-only-in-a-private-lan). + +You can also use `_is_parent`, `_is_child`, and any other host labels in both health entities and metrics +exporting. Speaking of which... + +### Host labels in alerts + +You can use host labels to logically organize your systems by their type, purpose, or location, and then apply specific +alerts to them. + +For example, let's use configuration example from earlier: + +```conf +[host labels] + type = webserver + location = us-seattle + installed = 20200218 +``` + +You could now create a new health entity (checking if disk space will run out soon) that applies only to any host +labeled `webserver`: + +```yaml + template: disk_fill_rate + on: disk.space + lookup: max -1s at -30m unaligned of avail + calc: ($this - $avail) / (30 * 60) + every: 15s + host labels: type = webserver +``` + +Or, by using one of the automatic labels, for only webserver systems running a specific OS: + +```yaml + host labels: _os_name = Debian* +``` + +In a streaming configuration where a parent node is triggering alerts for its child nodes, you could create health +entities that apply only to child nodes: + +```yaml + host labels: _is_child = true +``` + +Or when ephemeral Docker nodes are involved: + +```yaml + host labels: _container = docker +``` + +Of course, there are many more possibilities for intuitively organizing your systems with host labels. See the [health +documentation](/src/health/REFERENCE.md#alert-line-host-labels) for more details, and then get creative! + +### Host labels in metrics exporting + +If you have enabled any metrics exporting via our experimental [exporters](/src/exporting/README.md), any new host +labels you created manually are sent to the destination database alongside metrics. You can change this behavior by +editing `exporting.conf`, and you can even send automatically-generated labels on with exported metrics. + +```conf +[exporting:global] +enabled = yes +send configured labels = yes +send automatic labels = no +``` + +You can also change this behavior per exporting connection: + +```conf +[opentsdb:my_instance3] +enabled = yes +destination = localhost:4242 +data source = sum +update every = 10 +send charts matching = system.cpu +send configured labels = no +send automatic labels = yes +``` + +By applying labels to exported metrics, you can more easily parse historical metrics with the labels applied. To learn +more about exporting, read the [documentation](/src/exporting/README.md). + +## Metric labels + +The Netdata aggregate charts allow you to filter and group metrics based on label name-value pairs. + +All go.d plugin collectors support the specification of labels at the "collection job" level. Some collectors come with out of the box +labels (e.g. generic Prometheus collector, Kubernetes, Docker and more). But you can also add your own custom labels, by configuring +the data collection jobs. + +For example, suppose we have a single Netdata agent, collecting data from two remote Apache web servers, located in different data centers. +The web servers are load balanced and provide access to the service "Payments". + +You can define the following in `go.d.conf`, to be able to group the web requests by service or location: + +``` +jobs: + - name: mywebserver1 + url: http://host1/server-status?auto + labels: + service: "Payments" + location: "Atlanta" + - name: mywebserver2 + url: http://host2/server-status?auto + labels: + service: "Payments" + location: "New York" +``` + +Of course you may define as many custom label/value pairs as you like, in as many data collection jobs you need. diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/README.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/README.md new file mode 100644 index 000000000..00fe63af1 --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/README.md @@ -0,0 +1,34 @@ +# Running the Netdata Agent behind a reverse proxy + +If you need to access a Netdata agent's user interface or API in a production environment we recommend you put Netdata behind +another web server and secure access to the dashboard via SSL, user authentication and firewall rules. + +A dedicated web server also provides more robustness and capabilities than the Agent's [internal web server](/src/web/README.md). + +We have documented running behind +[nginx](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-nginx.md), +[Apache](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-apache.md), +[HAProxy](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-haproxy.md), +[Lighttpd](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-lighttpd.md), +[Caddy](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-caddy.md), +and [H2O](/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-h2o.md). +If you prefer a different web server, we suggest you follow the documentation for nginx and tell us how you did it + by adding your own "Running behind webserverX" document. + +When you run Netdata behind a reverse proxy, we recommend you firewall protect all your Netdata servers, so that only the web server IP will be allowed to directly access Netdata. To do this, run this on each of your servers (or use your firewall manager): + +```sh +PROXY_IP="1.2.3.4" +iptables -t filter -I INPUT -p tcp --dport 19999 \! -s ${PROXY_IP} -m conntrack --ctstate NEW -j DROP +``` + +The above will prevent anyone except your web server to access a Netdata dashboard running on the host. + +You can also use `netdata.conf`: + +``` +[web] + allow connections from = localhost 1.2.3.4 +``` + +Of course, you can add more IPs. diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-apache.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-apache.md new file mode 100644 index 000000000..1f7274d5c --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-apache.md @@ -0,0 +1,363 @@ +# Netdata via Apache's mod_proxy + +Below you can find instructions for configuring an apache server to: + +1. Proxy a single Netdata via an HTTP and HTTPS virtual host. +2. Dynamically proxy any number of Netdata servers. +3. Add user authentication. +4. Adjust Netdata settings to get optimal results. + +## Requirements + +Make sure your apache has `mod_proxy` and `mod_proxy_http` installed and enabled. + +On Debian/Ubuntu systems, install apache, which already includes the two modules, using: + +```sh +sudo apt-get install apache2 +``` + +Enable them: + +```sh +sudo a2enmod proxy +sudo a2enmod proxy_http +``` + +Also, enable the rewrite module: + +```sh +sudo a2enmod rewrite +``` +## Netdata on an existing virtual host + +On any **existing** and already **working** apache virtual host, you can redirect requests for URL `/netdata/` to one or more Netdata servers. + +### Proxy one Netdata, running on the same server apache runs + +Add the following on top of any existing virtual host. It will allow you to access Netdata as `http://virtual.host/netdata/`. + +```conf +<VirtualHost *:80> + + RewriteEngine On + ProxyRequests Off + ProxyPreserveHost On + + <Proxy *> + Require all granted + </Proxy> + + # Local Netdata server accessed with '/netdata/', at localhost:19999 + ProxyPass "/netdata/" "http://localhost:19999/" connectiontimeout=5 timeout=30 keepalive=on + ProxyPassReverse "/netdata/" "http://localhost:19999/" + + # if the user did not give the trailing /, add it + # for HTTP (if the virtualhost is HTTP, use this) + RewriteRule ^/netdata$ http://%{HTTP_HOST}/netdata/ [L,R=301] + # for HTTPS (if the virtualhost is HTTPS, use this) + #RewriteRule ^/netdata$ https://%{HTTP_HOST}/netdata/ [L,R=301] + + # rest of virtual host config here + +</VirtualHost> +``` + +### Proxy multiple Netdata running on multiple servers + +Add the following on top of any existing virtual host. It will allow you to access multiple Netdata as `http://virtual.host/netdata/HOSTNAME/`, where `HOSTNAME` is the hostname of any other Netdata server you have (to access the `localhost` Netdata, use `http://virtual.host/netdata/localhost/`). + +```conf +<VirtualHost *:80> + + RewriteEngine On + ProxyRequests Off + ProxyPreserveHost On + + <Proxy *> + Require all granted + </Proxy> + + # proxy any host, on port 19999 + ProxyPassMatch "^/netdata/([A-Za-z0-9\._-]+)/(.*)" "http://$1:19999/$2" connectiontimeout=5 timeout=30 keepalive=on + + # make sure the user did not forget to add a trailing / + # for HTTP (if the virtualhost is HTTP, use this) + RewriteRule "^/netdata/([A-Za-z0-9\._-]+)$" http://%{HTTP_HOST}/netdata/$1/ [L,R=301] + # for HTTPS (if the virtualhost is HTTPS, use this) + RewriteRule "^/netdata/([A-Za-z0-9\._-]+)$" https://%{HTTP_HOST}/netdata/$1/ [L,R=301] + + # rest of virtual host config here + +</VirtualHost> +``` + +> IMPORTANT<br/> +> The above config allows your apache users to connect to port 19999 on any server on your network. + +If you want to control the servers your users can connect to, replace the `ProxyPassMatch` line with the following. This allows only `server1`, `server2`, `server3` and `server4`. + +``` + ProxyPassMatch "^/netdata/(server1|server2|server3|server4)/(.*)" "http://$1:19999/$2" connectiontimeout=5 timeout=30 keepalive=on +``` + +## Netdata on a dedicated virtual host + +You can proxy Netdata through apache, using a dedicated apache virtual host. + +Create a new apache site: + +```sh +nano /etc/apache2/sites-available/netdata.conf +``` + +with this content: + +```conf +<VirtualHost *:80> + ProxyRequests Off + ProxyPreserveHost On + + ServerName netdata.domain.tld + + <Proxy *> + Require all granted + </Proxy> + + ProxyPass "/" "http://localhost:19999/" connectiontimeout=5 timeout=30 keepalive=on + ProxyPassReverse "/" "http://localhost:19999/" + + ErrorLog ${APACHE_LOG_DIR}/netdata-error.log + CustomLog ${APACHE_LOG_DIR}/netdata-access.log combined +</VirtualHost> +``` + +Enable the VirtualHost: + +```sh +sudo a2ensite netdata.conf && service apache2 reload +``` + +## Netdata proxy in Plesk + +_Assuming the main goal is to make Netdata running in HTTPS._ + +1. Make a subdomain for Netdata on which you enable and force HTTPS - You can use a free Let's Encrypt certificate +2. Go to "Apache & nginx Settings", and in the following section, add: + +```conf +RewriteEngine on +RewriteRule (.*) http://localhost:19999/$1 [P,L] +``` + +3. Optional: If your server is remote, then just replace "localhost" with your actual hostname or IP, it just works. + +Repeat the operation for as many servers as you need. + +## Enable Basic Auth + +If you wish to add an authentication (user/password) to access your Netdata, do these: + +Install the package `apache2-utils`. On Debian/Ubuntu run `sudo apt-get install apache2-utils`. + +Then, generate password for user `netdata`, using `htpasswd -c /etc/apache2/.htpasswd netdata` + +**Apache 2.2 Example:**\ +Modify the virtual host with these: + +```conf + # replace the <Proxy *> section + <Proxy *> + Order deny,allow + Allow from all + </Proxy> + + # add a <Location /netdata/> section + <Location /netdata/> + AuthType Basic + AuthName "Protected site" + AuthUserFile /etc/apache2/.htpasswd + Require valid-user + Order deny,allow + Allow from all + </Location> +``` + +Specify `Location /` if Netdata is running on dedicated virtual host. + +**Apache 2.4 (dedicated virtual host) Example:** + +```conf +<VirtualHost *:80> + RewriteEngine On + ProxyRequests Off + ProxyPreserveHost On + + ServerName netdata.domain.tld + + <Proxy *> + AllowOverride None + AuthType Basic + AuthName "Protected site" + AuthUserFile /etc/apache2/.htpasswd + Require valid-user + </Proxy> + + ProxyPass "/" "http://localhost:19999/" connectiontimeout=5 timeout=30 keepalive=on + ProxyPassReverse "/" "http://localhost:19999/" + + ErrorLog ${APACHE_LOG_DIR}/netdata-error.log + CustomLog ${APACHE_LOG_DIR}/netdata-access.log combined +</VirtualHost> +``` + +Note: Changes are applied by reloading or restarting Apache. + +## Configuration of Content Security Policy + +If you want to enable CSP within your Apache, you should consider some special requirements of the headers. Modify your configuration like that: + +``` + Header always set Content-Security-Policy "default-src http: 'unsafe-inline' 'self' 'unsafe-eval'; script-src http: 'unsafe-inline' 'self' 'unsafe-eval'; style-src http: 'self' 'unsafe-inline'" +``` + +Note: Changes are applied by reloading or restarting Apache. + +## Using Netdata with Apache's `mod_evasive` module + +The `mod_evasive` Apache module helps system administrators protect their web server from brute force and distributed +denial of service attack (DDoS) attacks. + +Because Netdata sends a request to the web server for every chart update, it's normal to create 20-30 requests per +second, per client. If you're using `mod_evasive` on your Apache web server, this volume of requests will trigger the +module's protection, and your dashboard will become unresponsive. You may even begin to see 403 errors. + +To mitigate this issue, you will need to change the value of the `DOSPageCount` option in your `mod_evasive.conf` file, +which can typically be found at `/etc/httpd/conf.d/mod_evasive.conf` or `/etc/apache2/mods-enabled/evasive.conf`. + +The `DOSPageCount` option sets the limit of the number of requests from a single IP address for the same page per page +interval, which is usually 1 second. The default value is `2` requests per second. Clearly, Netdata's typical usage will +exceed that threshold, and `mod_evasive` will add your IP address to a blocklist. + +Our users have found success by setting `DOSPageCount` to `30`. Try this, and raise the value if you continue to see 403 +errors while accessing the dashboard. + +```conf +DOSPageCount 30 +``` + +Restart Apache with `sudo systemctl restart apache2`, or the appropriate method to restart services on your system, to +reload its configuration with your new values. + +### Virtual host + +To adjust the `DOSPageCount` for a specific virtual host, open your virtual host config, which can be found at +`/etc/httpd/conf/sites-available/my-domain.conf` or `/etc/apache2/sites-available/my-domain.conf` and add the +following: + +```conf +<VirtualHost *:80> + ... + # Increase the DOSPageCount to prevent 403 errors and IP addresses being blocked. + <IfModule mod_evasive20.c> + DOSPageCount 30 + </IfModule> +</VirtualHost> +``` + +See issues [#2011](https://github.com/netdata/netdata/issues/2011) and +[#7658](https://github.com/netdata/netdata/issues/7568) for more information. + +# Netdata configuration + +You might edit `/etc/netdata/netdata.conf` to optimize your setup a bit. For applying these changes you need to restart Netdata. + +## Response compression + +If you plan to use Netdata exclusively via apache, you can gain some performance by preventing double compression of its output (Netdata compresses its response, apache re-compresses it) by editing `/etc/netdata/netdata.conf` and setting: + +``` +[web] + enable gzip compression = no +``` + +Once you disable compression at Netdata (and restart it), please verify you receive compressed responses from apache (it is important to receive compressed responses - the charts will be more snappy). + +## Limit direct access to Netdata + +You would also need to instruct Netdata to listen only on `localhost`, `127.0.0.1` or `::1`. + +``` +[web] + bind to = localhost +``` + +or + +``` +[web] + bind to = 127.0.0.1 +``` + +or + +``` +[web] + bind to = ::1 +``` + + + +You can also use a unix domain socket. This will also provide a faster route between apache and Netdata: + +``` +[web] + bind to = unix:/tmp/netdata.sock +``` + +Apache 2.4.24+ can not read from `/tmp` so create your socket in `/var/run/netdata` + +``` +[web] + bind to = unix:/var/run/netdata/netdata.sock +``` + +_note: Netdata v1.8+ support unix domain sockets_ + +At the apache side, prepend the 2nd argument to `ProxyPass` with `unix:/tmp/netdata.sock|`, like this: + +``` +ProxyPass "/netdata/" "unix:/tmp/netdata.sock|http://localhost:19999/" connectiontimeout=5 timeout=30 keepalive=on +``` + + + +If your apache server is not on localhost, you can set: + +``` +[web] + bind to = * + allow connections from = IP_OF_APACHE_SERVER +``` + +*note: Netdata v1.9+ support `allow connections from`* + +`allow connections from` accepts [Netdata simple patterns](/src/libnetdata/simple_pattern/README.md) to match against the connection IP address. + +## Prevent the double access.log + +apache logs accesses and Netdata logs them too. You can prevent Netdata from generating its access log, by setting this in `/etc/netdata/netdata.conf`: + +``` +[logs] + access = off +``` + +## Troubleshooting mod_proxy + +Make sure the requests reach Netdata, by examining `/var/log/netdata/access.log`. + +1. if the requests do not reach Netdata, your apache does not forward them. +2. if the requests reach Netdata but the URLs are wrong, you have not re-written them properly. + + diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-caddy.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-caddy.md new file mode 100644 index 000000000..b7608b309 --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-caddy.md @@ -0,0 +1,38 @@ +<!-- +title: "Netdata via Caddy" +custom_edit_url: "https://github.com/netdata/netdata/edit/master/docs/Running-behind-caddy.md" +sidebar_label: "Netdata via Caddy" +learn_status: "Published" +learn_topic_type: "Tasks" +learn_rel_path: "Configuration/Secure your nodes" +--> + +# Netdata via Caddy + +To run Netdata via [Caddy v2 proxying,](https://caddyserver.com/docs/caddyfile/directives/reverse_proxy) set your Caddyfile up like this: + +```caddyfile +netdata.domain.tld { + reverse_proxy localhost:19999 +} +``` + +Other directives can be added between the curly brackets as needed. + +To run Netdata in a subfolder: + +```caddyfile +netdata.domain.tld { + handle_path /netdata/* { + reverse_proxy localhost:19999 + } +} +``` + +## limit direct access to Netdata + +You would also need to instruct Netdata to listen only to `127.0.0.1` or `::1`. + +To limit access to Netdata only from localhost, set `bind socket to IP = 127.0.0.1` or `bind socket to IP = ::1` in `/etc/netdata/netdata.conf`. + + diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-h2o.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-h2o.md new file mode 100644 index 000000000..276b72e8b --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-h2o.md @@ -0,0 +1,187 @@ +<!-- +title: "Running Netdata behind H2O" +custom_edit_url: "https://github.com/netdata/netdata/edit/master/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-h2o.md" +sidebar_label: "Running Netdata behind H2O" +learn_status: "Published" +learn_topic_type: "Tasks" +learn_rel_path: "Configuration/Secure your nodes" +--> + +# Running Netdata behind H2O + +[H2O](https://h2o.examp1e.net/) is a new generation HTTP server that provides quicker response to users with less CPU utilization when compared to older generation of web servers. + +It is notable for having much simpler configuration than many popular HTTP servers, low resource requirements, and integrated native support for many things that other HTTP servers may need special setup to use. + +## Why H2O + +- Sane configuration defaults mean that typical configurations are very minimalistic and easy to work with. + +- Native support for HTTP/2 provides improved performance when accessing the Netdata dashboard remotely. + +- Password protect access to the Netdata dashboard without requiring Netdata Cloud. + +## H2O configuration file. + +On most systems, the H2O configuration is found under `/etc/h2o`. H2O uses [YAML 1.1](https://yaml.org/spec/1.1/), with a few special extensions, for it’s configuration files, with the main configuration file being `/etc/h2o/h2o.conf`. + +You can edit the H2O configuration file with Nano, Vim or any other text editors with which you are comfortable. + +After making changes to the configuration files, perform the following: + +- Test the configuration with `h2o -m test -c /etc/h2o/h2o.conf` + +- Restart H2O to apply tha changes with `/etc/init.d/h2o restart` or `service h2o restart` + +## Ways to access Netdata via H2O + +### As a virtual host + +With this method instead of `SERVER_IP_ADDRESS:19999`, the Netdata dashboard can be accessed via a human-readable URL such as `netdata.example.com` used in the configuration below. + +```yaml +hosts: + netdata.example.com: + listen: + port: 80 + paths: + /: + proxy.preserve-host: ON + proxy.reverse.url: http://127.0.0.1:19999 +``` + +### As a subfolder of an existing virtual host + +This method is recommended when Netdata is to be served from a subfolder (or directory). +In this case, the virtual host `netdata.example.com` already exists and Netdata has to be accessed via `netdata.example.com/netdata/`. + +```yaml +hosts: + netdata.example.com: + listen: + port: 80 + paths: + /netdata: + redirect: + status: 301 + url: /netdata/ + /netdata/: + proxy.preserve-host: ON + proxy.reverse.url: http://127.0.0.1:19999 +``` + +### As a subfolder for multiple Netdata servers, via one H2O instance + +This is the recommended configuration when one H2O instance will be used to manage multiple Netdata servers via subfolders. + +```yaml +hosts: + netdata.example.com: + listen: + port: 80 + paths: + /netdata/server1: + redirect: + status: 301 + url: /netdata/server1/ + /netdata/server1/: + proxy.preserve-host: ON + proxy.reverse.url: http://198.51.100.1:19999 + /netdata/server2: + redirect: + status: 301 + url: /netdata/server2/ + /netdata/server2/: + proxy.preserve-host: ON + proxy.reverse.url: http://198.51.100.2:19999 +``` + +Of course you can add as many backend servers as you like. + +Using the above, you access Netdata on the backend servers, like this: + +- `http://netdata.example.com/netdata/server1/` to reach Netdata on `198.51.100.1:19999` +- `http://netdata.example.com/netdata/server2/` to reach Netdata on `198.51.100.2:19999` + +### Encrypt the communication between H2O and Netdata + +In case Netdata's web server has been [configured to use TLS](/src/web/server/README.md#enabling-tls-support), it is +necessary to specify inside the H2O configuration that the final destination is using TLS. To do this, change the +`http://` on the `proxy.reverse.url` line in your H2O configuration with `https://` + +### Enable authentication + +Create an authentication file to enable basic authentication via H2O, this secures your Netdata dashboard. + +If you don't have an authentication file, you can use the following command: + +```sh +printf "yourusername:$(openssl passwd -apr1)" > /etc/h2o/passwords +``` + +And then add a basic authentication handler to each path definition: + +```yaml +hosts: + netdata.example.com: + listen: + port: 80 + paths: + /: + mruby.handler: | + require "htpasswd.rb" + Htpasswd.new("/etc/h2o/passwords", "netdata.example.com") + proxy.preserve-host: ON + proxy.reverse.url: http://127.0.0.1:19999 +``` + +For more information on using basic authentication with H2O, see [their official documentation](https://h2o.examp1e.net/configure/basic_auth.html). + +## Limit direct access to Netdata + +If your H2O server is on `localhost`, you can use this to ensure external access is only possible through H2O: + +``` +[web] + bind to = 127.0.0.1 ::1 +``` + + + +You can also use a unix domain socket. This will provide faster communication between H2O and Netdata as well: + +``` +[web] + bind to = unix:/run/netdata/netdata.sock +``` + +In the H2O configuration, use a line like the following to connect to Netdata via the unix socket: + +```yaml +proxy.reverse.url http://[unix:/run/netdata/netdata.sock] +``` + + + +If your H2O server is not on localhost, you can set: + +``` +[web] + bind to = * + allow connections from = IP_OF_H2O_SERVER +``` + +*note: Netdata v1.9+ support `allow connections from`* + +`allow connections from` accepts [Netdata simple patterns](/src/libnetdata/simple_pattern/README.md) to match against +the connection IP address. + +## Prevent the double access.log + +H2O logs accesses and Netdata logs them too. You can prevent Netdata from generating its access log, by setting +this in `/etc/netdata/netdata.conf`: + +``` +[logs] + access = off +``` diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-haproxy.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-haproxy.md new file mode 100644 index 000000000..9d2aff670 --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-haproxy.md @@ -0,0 +1,297 @@ +<!-- +title: "Netdata via HAProxy" +custom_edit_url: "https://github.com/netdata/netdata/edit/master/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-haproxy.md" +sidebar_label: "Netdata via HAProxy" +learn_status: "Published" +learn_topic_type: "Tasks" +learn_rel_path: "Configuration/Secure your nodes" +--> + +# Netdata via HAProxy + +> HAProxy is a free, very fast and reliable solution offering high availability, load balancing, +> and proxying for TCP and HTTP-based applications. It is particularly suited for very high traffic websites +> and powers quite a number of the world's most visited ones. + +If Netdata is running on a host running HAProxy, rather than connecting to Netdata from a port number, a domain name can +be pointed at HAProxy, and HAProxy can redirect connections to the Netdata port. This can make it possible to connect to +Netdata at `https://example.com` or `https://example.com/netdata/`, which is a much nicer experience then +`http://example.com:19999`. + +To proxy requests from [HAProxy](https://github.com/haproxy/haproxy) to Netdata, +the following configuration can be used: + +## Default Configuration + +For all examples, set the mode to `http` + +```conf +defaults + mode http +``` + +## Simple Configuration + +A simple example where the base URL, say `http://example.com`, is used with no subpath: + +### Frontend + +Create a frontend to receive the request. + +```conf +frontend http_frontend + ## HTTP ipv4 and ipv6 on all ips ## + bind :::80 v4v6 + + default_backend netdata_backend +``` + +### Backend + +Create the Netdata backend which will send requests to port `19999`. + +```conf +backend netdata_backend + option forwardfor + server netdata_local 127.0.0.1:19999 + + http-request set-header Host %[src] + http-request set-header X-Forwarded-For %[src] + http-request set-header X-Forwarded-Port %[dst_port] + http-request set-header Connection "keep-alive" +``` + +## Configuration with subpath + +An example where the base URL is used with a subpath `/netdata/`: + +### Frontend + +To use a subpath, create an ACL, which will set a variable based on the subpath. + +```conf +frontend http_frontend + ## HTTP ipv4 and ipv6 on all ips ## + bind :::80 v4v6 + + # URL begins with /netdata + acl is_netdata url_beg /netdata + + # if trailing slash is missing, redirect to /netdata/ + http-request redirect scheme https drop-query append-slash if is_netdata ! { path_beg /netdata/ } + + ## Backends ## + use_backend netdata_backend if is_netdata + + # Other requests go here (optional) + # put netdata_backend here if no others are used + default_backend www_backend +``` + +### Backend + +Same as simple example, except remove `/netdata/` with regex. + +```conf +backend netdata_backend + option forwardfor + server netdata_local 127.0.0.1:19999 + + http-request set-path %[path,regsub(^/netdata/,/)] + + http-request set-header Host %[src] + http-request set-header X-Forwarded-For %[src] + http-request set-header X-Forwarded-Port %[dst_port] + http-request set-header Connection "keep-alive" +``` + +## Using TLS communication + +TLS can be used by adding port `443` and a cert to the frontend. +This example will only use Netdata if host matches example.com (replace with your domain). + +### Frontend + +This frontend uses a certificate list. + +```conf +frontend https_frontend + ## HTTP ## + bind :::80 v4v6 + # Redirect all HTTP traffic to HTTPS with 301 redirect + redirect scheme https code 301 if !{ ssl_fc } + + ## HTTPS ## + # Bind to all v4/v6 addresses, use a list of certs in file + bind :::443 v4v6 ssl crt-list /etc/letsencrypt/certslist.txt + + ## ACL ## + # Optionally check host for Netdata + acl is_example_host hdr_sub(host) -i example.com + + ## Backends ## + use_backend netdata_backend if is_example_host + # Other requests go here (optional) + default_backend www_backend +``` + +In the cert list file place a mapping from a certificate file to the domain used: + +`/etc/letsencrypt/certslist.txt`: + +```txt +example.com /etc/letsencrypt/live/example.com/example.com.pem +``` + +The file `/etc/letsencrypt/live/example.com/example.com.pem` should contain the key and +certificate (in that order) concatenated into a `.pem` file.: + +```sh +cat /etc/letsencrypt/live/example.com/fullchain.pem \ + /etc/letsencrypt/live/example.com/privkey.pem > \ + /etc/letsencrypt/live/example.com/example.com.pem +``` + +### Backend + +Same as simple, except set protocol `https`. + +```conf +backend netdata_backend + option forwardfor + server netdata_local 127.0.0.1:19999 + + http-request add-header X-Forwarded-Proto https + http-request set-header Host %[src] + http-request set-header X-Forwarded-For %[src] + http-request set-header X-Forwarded-Port %[dst_port] + http-request set-header Connection "keep-alive" +``` + +## Enable authentication + +To use basic HTTP Authentication, create an authentication list: + +```conf +# HTTP Auth +userlist basic-auth-list + group is-admin + # Plaintext password + user admin password passwordhere groups is-admin +``` + +You can create a hashed password using the `mkpassword` utility. + +```sh + printf "passwordhere" | mkpasswd --stdin --method=sha-256 +$5$l7Gk0VPIpKO$f5iEcxvjfdF11khw.utzSKqP7W.0oq8wX9nJwPLwzy1 +``` + +Replace `passwordhere` with hash: + +```conf +user admin password $5$l7Gk0VPIpKO$f5iEcxvjfdF11khw.utzSKqP7W.0oq8wX9nJwPLwzy1 groups is-admin +``` + +Now add at the top of the backend: + +```conf +acl devops-auth http_auth_group(basic-auth-list) is-admin +http-request auth realm netdata_local unless devops-auth +``` + +## Full Example + +Full example configuration with HTTP auth over TLS with subpath: + +```conf +global + maxconn 20000 + + log /dev/log local0 + log /dev/log local1 notice + user haproxy + group haproxy + pidfile /run/haproxy.pid + + stats socket /run/haproxy/admin.sock mode 660 level admin expose-fd listeners + stats timeout 30s + daemon + + tune.ssl.default-dh-param 4096 # Max size of DHE key + + # Default ciphers to use on SSL-enabled listening sockets. + ssl-default-bind-ciphers ECDH+AESGCM:DH+AESGCM:ECDH+AES256:DH+AES256:ECDH+AES128:DH+AES:RSA+AESGCM:RSA+AES:!aNULL:!MD5:!DSS + ssl-default-bind-options no-sslv3 + +defaults + log global + mode http + option httplog + option dontlognull + timeout connect 5000 + timeout client 50000 + timeout server 50000 + errorfile 400 /etc/haproxy/errors/400.http + errorfile 403 /etc/haproxy/errors/403.http + errorfile 408 /etc/haproxy/errors/408.http + errorfile 500 /etc/haproxy/errors/500.http + errorfile 502 /etc/haproxy/errors/502.http + errorfile 503 /etc/haproxy/errors/503.http + errorfile 504 /etc/haproxy/errors/504.http + +frontend https_frontend + ## HTTP ## + bind :::80 v4v6 + # Redirect all HTTP traffic to HTTPS with 301 redirect + redirect scheme https code 301 if !{ ssl_fc } + + ## HTTPS ## + # Bind to all v4/v6 addresses, use a list of certs in file + bind :::443 v4v6 ssl crt-list /etc/letsencrypt/certslist.txt + + ## ACL ## + # Optionally check host for Netdata + acl is_example_host hdr_sub(host) -i example.com + acl is_netdata url_beg /netdata + + http-request redirect scheme https drop-query append-slash if is_netdata ! { path_beg /netdata/ } + + ## Backends ## + use_backend netdata_backend if is_example_host is_netdata + default_backend www_backend + +# HTTP Auth +userlist basic-auth-list + group is-admin + # Hashed password + user admin password $5$l7Gk0VPIpKO$f5iEcxvjfdF11khw.utzSKqP7W.0oq8wX9nJwPLwzy1 groups is-admin + +## Default server(s) (optional)## +backend www_backend + mode http + balance roundrobin + timeout connect 5s + timeout server 30s + timeout queue 30s + + http-request add-header 'X-Forwarded-Proto: https' + server other_server 111.111.111.111:80 check + +backend netdata_backend + acl devops-auth http_auth_group(basic-auth-list) is-admin + http-request auth realm netdata_local unless devops-auth + + option forwardfor + server netdata_local 127.0.0.1:19999 + + http-request set-path %[path,regsub(^/netdata/,/)] + + http-request add-header X-Forwarded-Proto https + http-request set-header Host %[src] + http-request set-header X-Forwarded-For %[src] + http-request set-header X-Forwarded-Port %[dst_port] + http-request set-header Connection "keep-alive" +``` + + diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-lighttpd.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-lighttpd.md new file mode 100644 index 000000000..637bc0642 --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-lighttpd.md @@ -0,0 +1,75 @@ +<!-- +title: "Netdata via lighttpd v1.4.x" +custom_edit_url: "https://github.com/netdata/netdata/edit/master/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-lighttpd.md" +sidebar_label: "Netdata via lighttpd v1.4.x" +learn_status: "Published" +learn_topic_type: "Tasks" +learn_rel_path: "Configuration/Secure your nodes" +--> + +# Netdata via lighttpd v1.4.x + +Here is a config for accessing Netdata in a suburl via lighttpd 1.4.46 and newer: + +```txt +$HTTP["url"] =~ "^/netdata/" { + proxy.server = ( "" => ("netdata" => ( "host" => "127.0.0.1", "port" => 19999 ))) + proxy.header = ( "map-urlpath" => ( "/netdata/" => "/") ) +} +``` + +If you have older lighttpd you have to use a chain (such as below), as explained [at this stackoverflow answer](http://stackoverflow.com/questions/14536554/lighttpd-configuration-to-proxy-rewrite-from-one-domain-to-another). + +```txt +$HTTP["url"] =~ "^/netdata/" { + proxy.server = ( "" => ("" => ( "host" => "127.0.0.1", "port" => 19998 ))) +} + +$SERVER["socket"] == ":19998" { + url.rewrite-once = ( "^/netdata(.*)$" => "/$1" ) + proxy.server = ( "" => ( "" => ( "host" => "127.0.0.1", "port" => 19999 ))) +} +``` + + + +If the only thing the server is exposing via the web is Netdata (and thus no suburl rewriting required), +then you can get away with just + +``` +proxy.server = ( "" => ( ( "host" => "127.0.0.1", "port" => 19999 ))) +``` + +Though if it's public facing you might then want to put some authentication on it. htdigest support +looks like: + +``` +auth.backend = "htdigest" +auth.backend.htdigest.userfile = "/etc/lighttpd/lighttpd.htdigest" +auth.require = ( "" => ( "method" => "digest", + "realm" => "netdata", + "require" => "valid-user" + ) + ) +``` + +other auth methods, and more info on htdigest, can be found in lighttpd's [mod_auth docs](http://redmine.lighttpd.net/projects/lighttpd/wiki/Docs_ModAuth). + + + +It seems that lighttpd (or some versions of it), fail to proxy compressed web responses. +To solve this issue, disable web response compression in Netdata. + +Open `/etc/netdata/netdata.conf` and set in [global]\: + +``` +enable web responses gzip compression = no +``` + +## limit direct access to Netdata + +You would also need to instruct Netdata to listen only to `127.0.0.1` or `::1`. + +To limit access to Netdata only from localhost, set `bind socket to IP = 127.0.0.1` or `bind socket to IP = ::1` in `/etc/netdata/netdata.conf`. + + diff --git a/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-nginx.md b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-nginx.md new file mode 100644 index 000000000..f2dd137dd --- /dev/null +++ b/docs/netdata-agent/configuration/running-the-netdata-agent-behind-a-reverse-proxy/Running-behind-nginx.md @@ -0,0 +1,282 @@ +# Running Netdata behind Nginx + +## Intro + +[Nginx](https://nginx.org/en/) is an HTTP and reverse proxy server, a mail proxy server, and a generic TCP/UDP proxy server used to host websites and applications of all sizes. + +The software is known for its low impact on memory resources, high scalability, and its modular, event-driven architecture which can offer secure, predictable performance. + +## Why Nginx + +- By default, Nginx is fast and lightweight out of the box. + +- Nginx is used and useful in cases when you want to access different instances of Netdata from a single server. + +- Password-protect access to Netdata, until distributed authentication is implemented via the Netdata cloud Sign In mechanism. + +- A proxy was necessary to encrypt the communication to Netdata, until v1.16.0, which provided TLS (HTTPS) support. + +## Nginx configuration file + +All Nginx configurations can be found in the `/etc/nginx/` directory. The main configuration file is `/etc/nginx/nginx.conf`. Website or app-specific configurations can be found in the `/etc/nginx/site-available/` directory. + +Configuration options in Nginx are known as directives. Directives are organized into groups known as blocks or contexts. The two terms can be used interchangeably. + +Depending on your installation source, you’ll find an example configuration file at `/etc/nginx/conf.d/default.conf` or `etc/nginx/sites-enabled/default`, in some cases you may have to manually create the `sites-available` and `sites-enabled` directories. + +You can edit the Nginx configuration file with Nano, Vim or any other text editors you are comfortable with. + +After making changes to the configuration files: + +- Test Nginx configuration with `nginx -t`. + +- Restart Nginx to effect the change with `/etc/init.d/nginx restart` or `service nginx restart`. + +## Ways to access Netdata via Nginx + +### As a virtual host + +With this method instead of `SERVER_IP_ADDRESS:19999`, the Netdata dashboard can be accessed via a human-readable URL such as `netdata.example.com` used in the configuration below. + +```conf +upstream backend { + # the Netdata server + server 127.0.0.1:19999; + keepalive 1024; +} + +server { + # nginx listens to this + listen 80; + # uncomment the line if you want nginx to listen on IPv6 address + #listen [::]:80; + + # the virtual host name of this + server_name netdata.example.com; + + location / { + proxy_set_header X-Forwarded-Host $host; + proxy_set_header X-Forwarded-Server $host; + proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; + proxy_pass http://backend; + proxy_http_version 1.1; + proxy_pass_request_headers on; + proxy_set_header Connection "keep-alive"; + proxy_store off; + } +} +``` + +### As a subfolder to an existing virtual host + +This method is recommended when Netdata is to be served from a subfolder (or directory). +In this case, the virtual host `netdata.example.com` already exists and Netdata has to be accessed via `netdata.example.com/netdata/`. + +```conf +upstream netdata { + server 127.0.0.1:19999; + keepalive 64; +} + +server { + listen 80; + # uncomment the line if you want nginx to listen on IPv6 address + #listen [::]:80; + + # the virtual host name of this subfolder should be exposed + #server_name netdata.example.com; + + location = /netdata { + return 301 /netdata/; + } + + location ~ /netdata/(?<ndpath>.*) { + proxy_redirect off; + proxy_set_header Host $host; + + proxy_set_header X-Forwarded-Host $host; + proxy_set_header X-Forwarded-Server $host; + proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; + proxy_http_version 1.1; + proxy_pass_request_headers on; + proxy_set_header Connection "keep-alive"; + proxy_store off; + proxy_pass http://netdata/$ndpath$is_args$args; + + gzip on; + gzip_proxied any; + gzip_types *; + } +} +``` + +### As a subfolder for multiple Netdata servers, via one Nginx + +This is the recommended configuration when one Nginx will be used to manage multiple Netdata servers via subfolders. + +```conf +upstream backend-server1 { + server 10.1.1.103:19999; + keepalive 64; +} +upstream backend-server2 { + server 10.1.1.104:19999; + keepalive 64; +} + +server { + listen 80; + # uncomment the line if you want nginx to listen on IPv6 address + #listen [::]:80; + + # the virtual host name of this subfolder should be exposed + #server_name netdata.example.com; + + location ~ /netdata/(?<behost>.*?)/(?<ndpath>.*) { + proxy_set_header X-Forwarded-Host $host; + proxy_set_header X-Forwarded-Server $host; + proxy_set_header X-Forwarded-For $proxy_add_x_forwarded_for; + proxy_http_version 1.1; + proxy_pass_request_headers on; + proxy_set_header Connection "keep-alive"; + proxy_store off; + proxy_pass http://backend-$behost/$ndpath$is_args$args; + + gzip on; + gzip_proxied any; + gzip_types *; + } + + # make sure there is a trailing slash at the browser + # or the URLs will be wrong + location ~ /netdata/(?<behost>.*) { + return 301 /netdata/$behost/; + } +} +``` + +Of course you can add as many backend servers as you like. + +Using the above, you access Netdata on the backend servers, like this: + +- `http://netdata.example.com/netdata/server1/` to reach `backend-server1` +- `http://netdata.example.com/netdata/server2/` to reach `backend-server2` + +### Encrypt the communication between Nginx and Netdata + +In case Netdata's web server has been [configured to use TLS](/src/web/server/README.md#enabling-tls-support), it is +necessary to specify inside the Nginx configuration that the final destination is using TLS. To do this, please, append +the following parameters in your `nginx.conf` + +```conf +proxy_set_header X-Forwarded-Proto https; +proxy_pass https://localhost:19999; +``` + +Optionally it is also possible to [enable TLS/SSL on Nginx](http://nginx.org/en/docs/http/configuring_https_servers.html), this way the user will encrypt not only the communication between Nginx and Netdata but also between the user and Nginx. + +If Nginx is not configured as described here, you will probably receive the error `SSL_ERROR_RX_RECORD_TOO_LONG`. + +### Enable authentication + +Create an authentication file to enable basic authentication via Nginx, this secures your Netdata dashboard. + +If you don't have an authentication file, you can use the following command: + +```sh +printf "yourusername:$(openssl passwd -apr1)" > /etc/nginx/passwords +``` + +And then enable the authentication inside your server directive: + +```conf +server { + # ... + auth_basic "Protected"; + auth_basic_user_file passwords; + # ... +} +``` + +## Limit direct access to Netdata + +If your Nginx is on `localhost`, you can use this to protect your Netdata: + +``` +[web] + bind to = 127.0.0.1 ::1 +``` + +You can also use a unix domain socket. This will also provide a faster route between Nginx and Netdata: + +``` +[web] + bind to = unix:/var/run/netdata/netdata.sock +``` + +*note: Netdata v1.8+ support unix domain sockets* + +At the Nginx side, use something like this to use the same unix domain socket: + +```conf +upstream backend { + server unix:/var/run/netdata/netdata.sock; + keepalive 64; +} +``` + + +If your Nginx server is not on localhost, you can set: + +``` +[web] + bind to = * + allow connections from = IP_OF_NGINX_SERVER +``` + +*note: Netdata v1.9+ support `allow connections from`* + +`allow connections from` accepts [Netdata simple patterns](/src/libnetdata/simple_pattern/README.md) to match against the +connection IP address. + +## Prevent the double access.log + +Nginx logs accesses and Netdata logs them too. You can prevent Netdata from generating its access log, by setting this in `/etc/netdata/netdata.conf`: + +``` +[logs] + access = off +``` + +## Use gzip compression + +By default, netdata compresses its responses. You can have nginx do that instead, with the following options in the `location /` block: + +```conf + location / { + ... + gzip on; + gzip_proxied any; + gzip_types *; + } +``` + +To disable Netdata's gzip compression, open `netdata.conf` and in the `[web]` section put: + +```conf +[web] + enable gzip compression = no +``` + +## SELinux + +If you get an 502 Bad Gateway error you might check your Nginx error log: + +```sh +# cat /var/log/nginx/error.log: +2016/09/09 12:34:05 [crit] 5731#5731: *1 connect() to 127.0.0.1:19999 failed (13: Permission denied) while connecting to upstream, client: 1.2.3.4, server: netdata.example.com, request: "GET / HTTP/2.0", upstream: "http://127.0.0.1:19999/", host: "netdata.example.com" +``` + +If you see something like the above, chances are high that SELinux prevents nginx from connecting to the backend server. To fix that, just use this policy: `setsebool -P httpd_can_network_connect true`. + + |