LogDNA Agent streams from log files to your LogDNA account. Works with Linux, Windows, and macOS Servers.
logdna-agent will soon be deprecated and will cease to have support. Please refer to LogDNA Agent V2 for all future implementations.
Check out the official LogDNA site to learn how to sign up for an account and get started.
Follow these instructions to run the LogDNA agent from source:
git clone https://github.com/logdna/logdna-agent.git
cd logdna-agent
npm install
# help
sudo node index.js --help
# configure
sudo node index.js -k <YOUR LOGDNA INGESTION KEY>
# On Linux, this will generate a config file at: /etc/logdna.conf
# On Windows, this will generate a config file at: C:\ProgramData\logdna\logdna.conf
# on Linux, /var/log is monitored/added by default (recursively). You can optionally specify more folders
# on Windows, C:\ProgramData\logs is monitored/added by default (recursively). You can optionally specify more folders
sudo node index.js -d /path/to/log/folders -d /path/to/2nd/folder
sudo node index.js -d /var/log # folder only assumes *.log + extensionless files
sudo node index.js -d "/var/log/*.txt" # supports glob patterns
sudo node index.js -d "/var/log/**/*.txt" # *.txt in any subfolder
sudo node index.js -d "/var/log/**/myapp.log" # myapp.log in any subfolder
sudo node index.js -d "/var/log/+(name1|name2).log" # supports extended glob patterns
sudo node index.js -e /var/log/nginx/error.log # exclude specific files from -d
sudo node index.js -f /usr/local/nginx/logs/access.log # add specific files
sudo node index.js -t production # tags
sudo node index.js -t production,app1=/opt/app1 # tags for specific paths
# other commands
sudo node index.js -l # show all saved options from config
sudo node index.js -l tags,key,logdir # show specific entries from config
sudo node index.js -u tags # unset tags
sudo node index.js -u tags,logdir # unset tags and logdir
sudo node index.js -u all # unset everything except ingestion key
# start the agent
sudo node index.jsNote that when using glob patterns with index.js, you must enclose the pattern in double quotes.
Normally a config file is automatically generated (e.g. when you set a key using index.js -k) and updated (e.g. when you add a directory
using index.js -d) but you can create your own config file /etc/logdna.conf on Linux and C:\ProgramData\logdna\logdna.conf on Windows and
save your settings there:
logdir = /var/log/myapp,/path/to/2nd/dir
key = <YOUR LOGDNA INGESTION KEY>
On Windows, use \\ as a separator:
logdir = C:\\Users\\username\\AppData\\myapp
key = <YOUR LOGDNA INGESTION KEY>
logdir: sets the paths that the agent will monitor for new files. Multiple paths can be specified, separated by,. Supports glob patterns + specific files. By default this option is set to monitor.logand extensionless files under/var/log/.exclude: sets files to exclude that would otherwise match what's set inlogdir. Multiple paths can be specified, separated by,. Supports glob patterns + specific filesexclude_regex: filters out any log lines matching this pattern in any file. Should not include leading or trailing/.key: your LogDNA Ingestion Key. You can obtain one by creating an account at LogDNA. Once logged in, click on the Gear icon, then Account Profile to find your key.tags: tags can be used e.g. to separate data from production, staging, or autoscaling use caseshostname: set this to override the os hostname
- The Agent maintains persistent connections to LogDNA ingestion servers with HTTPS encryption
- Reconnects if disconnected and will queue up new log lines while disconnected
- Compresses on upload (gzip)
- Rescans for new files in all
logdirpaths every minute - Handles log rotated files on most operating systems (supports: renamed, truncated & "new file per day" log rotation methods)
- Init script is available here (rpm/deb packages already include this)
# YUM Repo
echo "[logdna]
name=LogDNA packages
baseurl=https://repo.logdna.com/el6/
enabled=1
gpgcheck=1
gpgkey=https://repo.logdna.com/logdna.gpg" | sudo tee /etc/yum.repos.d/logdna.repo
# APT Repo
echo "deb https://repo.logdna.com stable main" | sudo tee /etc/apt/sources.list.d/logdna.list
wget -O- https://repo.logdna.com/logdna.gpg | sudo apt-key add -
sudo apt-get update
The LogDNA agent authenticates using your LogDNA Ingestion Key and opens a secure web socket to LogDNA's ingestion servers. It then 'tails' for new log data, as well as watches for new files added to your specific logging directories.
If you don't have a LogDNA account, you can create one on https://logdna.com. Or if you're using macOS w/Homebrew installed:
brew cask install logdna-cli
logdna register <email>
Please see our documentation for Kubernetes instructions.
OpenShift logging requires a few additional steps over Kubernetes, but still pretty easy! Like Kubernetes, we extract pertinent metadata: pod name, container name, container id, namespace, project, and labels etc:
oc adm new-project --node-selector='' logdna-agent
oc project logdna-agent
oc create serviceaccount logdna-agent
oc adm policy add-scc-to-user privileged system:serviceaccount:logdna-agent:logdna-agent
oc create secret generic logdna-agent-key --from-literal=logdna-agent-key=<YOUR LOGDNA INGESTION KEY>
oc create -f https://raw.githubusercontent.com/logdna/logdna-agent/master/logdna-agent-ds-os.yaml
This automatically installs a logdna-agent pod into each node in your cluster and ships stdout/stderr from all containers, both application logs
and node logs. Note: By default, the agent pod will collect logs from all namespaces on each node, including kube-system. View your logs
at https://app.logdna.com. See YAML file for additional
options such as LOGDNA_TAGS.
Notes:
- The
oc adm new-projectmethod prevents having to adjust the project namespace's node-selector after creation. - This uses
JOURNALD=files, you may need to change this if you have changed OpenShift logging configuration. - This has been tested on OpenShift 3.5-11
If you're using a docker-less containerized (e.g. podman) operating system like Fedora CoreOS where logs are forwarded to journald, you will be unable to use logspout (if your OS does use Docker, refer to LogDNA LogSpout for instructions on how to set that up).
You can run logdna-agent inside a container to read from journald with a few
modifications. First, you'll need to set up systemd inside the container so
that it can read from journalctl. Note that due to the systemd dependency,
you may have some difficulties starting from a base image like Alpine. Instead,
try starting from a distribution such as Debian or CentOS. The following
assumes that you're starting from the Ubuntu based image provided by LogDNA.
In your Containerfile, install systemd:
FROM logdna/logdna-agent
# Install systemd so we can read logs via journalctl
RUN apt-get update \
&& apt-get install -y --no-install-recommends systemd \
&& rm -rf /var/lib/apt/lists/*
Next, you need to ensure the USEJOURNALD environment variable is set. If set
to files, the agent will read from journald and forward the logs. The agent
can be configured either in your image or entrypoint.
By default the agent will read logs from the /var/log. These logs won't be
very useful since they'll be referencing the container and not the host. You
cannot omit logdir since the agent will still read from /var/log if
logdir is missing from the configuration. Instead, to disable reading from
files, set logdir to an empty directory (e.g. /usr/src).
Finally, mount /var/log/journal inside the container so the agent can use
journalctl to read logs. An example systemd service configuration could look
like the following:
[Unit]
Description=LogDNA Forwarder
After=network-online.target
Wants=network-online.target
[Service]
Restart=on-failure
ExecStartPre=-/bin/podman kill logdna
ExecStartPre=-/bin/podman rm logdna
ExecStartPre=/bin/podman pull my-custom/logdna-agent
ExecStart=/bin/podman run -v /var/log/journal:/var/log/journal:z --name logdna my-custom/logdna-agent
[Install]
WantedBy=multi-user.target
The LogDNA agent can be installed through Chocolatey. You will need:
- Windows PowerShell v3+ (not PowerShell Core aka PowerShell 6 yet)
- Windows 7+/Windows 2003+ (Server Core also, but not Windows Nano Server)
- .NET Framework 4.x+
For more details, view our Windows Agent docs
Our paid plans start at $1.50/GB per month. Pay only for what you use, with no fixed data buckets. All paid plans include all features.
- Lee Liu (LogDNA)
- Mike Hu (LogDNA)
- Ryan Staatz (LogDNA)
- Samir Musali (LogDNA)
- Steven Edouard (Microsoft)
- Rita Zhang (Microsoft)
- Lauren Tran (Microsoft)
Contributions are always welcome. See the contributing guide to learn how you can help. Build instructions for the agent are also in the guide.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent
