Vector by datahog - Vendor neutral log pipeline manager
👤 SUVANKAR SARKAR •
📅 April 5, 2026 •
👁️ 23 views
• 🔄 Updated April 15, 2026
Observability
log
data pipe
Capable to collect transform and push to multiple destination, Vendor neutral and lock-in free Built in **[Rust](https://rust-lang.org/)** so memory efficient and safe. ***Single binary X86\_64, ARM64/v7 \**No runtime**
```
curl --proto '=https' --tlsv1.2 -sSfL https://sh.vector.dev | bash
wget https://packages.timber.io/vector/0.38.0/vector-amd64.deb
bash -c "$(curl -L https://setup.vector.dev)" && apt-get install vector #
vector validate /etc/vector/vector.yaml # /root/.vector/config/vector.yaml
vector test /etc/vector/vector.yaml
# Dry run (process but don't send)
vector --config /etc/vector/vector.yaml --dry-run
# Check Vector metrics
curl http://localhost:9598/metrics
```
**/etc/vector/vector.yaml**
```
sources:
vector_metrics:
type: internal_metrics
scrape_interval_secs: 15
# Host metrics (CPU, memory, disk, network)
host_metrics:
type: host_metrics
collectors:
- cpu
- disk
- filesystem
- load
- host
- memory
- network
scrape_interval_secs: 15
# Systemd service metrics
systemd_metrics:
type: systemd_metrics
scrape_interval_secs: 30
# Process metrics (like ps/top)
process_metrics:
type: internal_metrics
scrape_interval_secs: 15
syslog:
type: file
include:
- "/var/log/syslog"
- "/var/log/auth.log"
read_from: beginning
nginx:
type: file
include:
- "/var/log/nginx/*.log"
read_from: beginning
transforms:
# Add custom tags/labels
add_tags:
type: add_fields
inputs: ["host_metrics", "systemd_metrics", "process_metrics"]
fields:
environment: "production"
region: "us-east-1"
collector: "vector"
sinks:
loki:
type: loki
inputs: ["syslog", "nginx"]
endpoint: "http://your-loki-server:3100"
labels:
host: "{{ host }}"
job: "vector"
encoding:
codec: json
tls:
ca_file: /etc/vector/certs/ca.crt
crt_file: /etc/vector/certs/vector.crt
key_file: /etc/vector/certs/vector.key
verify_certificate: true
verify_hostname: true
elasticsearch:
type: elasticsearch
inputs: ["app_logs"]
endpoint: "http://elasticsearch:9200"
index: "app-logs-%Y-%m-%d"
s3:
type: aws_s3
inputs: ["app_logs"]
bucket: "my-logs-bucket"
key_prefix: "logs/"
# Optional: Also output to console for debugging
console:
type: console
inputs: ["syslog"]
encoding:
codec: json
# Vector's GraphQL API (disabled by default)
# Uncomment to try it out with the `vector top` command or
# in your browser at http://localhost:8686
# api:
# enabled: true
# address: "127.0.0.1:8686"
```
**Use Systemd to manage vector**
```
# Vector is automatically installed as a systemd service
systemctl start vector
systemctl enable vector
systemctl status vector
# View logs
journalctl -u vector -f
```
**Configuration Directory**
```
# Main config
/etc/vector/vector.yaml
# Config directory (for multiple config files)
/etc/vector/vector.d/
# Data directory
/var/lib/vector/
# Logs
journalctl -u vector
```