Linux Integration
System-level CPU, memory, disk, network, and process metrics via the TigerOps agent. Monitor your entire Linux fleet with AI capacity forecasting and kernel event correlation.
How It Works
Install TigerOps Agent
Install the TigerOps agent on your Linux hosts using apt, yum, or the one-line install script. The agent runs as a systemd service and begins collecting metrics within 30 seconds.
Auto-Discover Host Metrics
The TigerOps agent automatically discovers and collects CPU, memory, disk, network, load average, file descriptors, and running process metrics without manual configuration.
Enable Log Collection
Configure the agent to tail syslog, auth.log, kern.log, and application log files. TigerOps correlates log events with metric spikes for unified host observability.
Scale Across Your Fleet
Deploy the agent across your Linux fleet via Ansible, Chef, Puppet, or Terraform. Tag hosts with environment, region, and role labels for fleet-wide dashboards and alerting.
What You Get Out of the Box
CPU and Load Average Monitoring
Per-core CPU utilization (user, system, iowait, steal, irq), load average (1m, 5m, 15m), context switches, and CPU frequency scaling. TigerOps detects CPU steal on noisy-neighbor cloud instances.
Memory and Swap Tracking
Physical memory (total, used, free, cached, buffered, available), swap utilization, and slab cache breakdown. AI alerts on memory pressure trends days before OOM events occur.
Disk I/O and Filesystem Metrics
Per-device read/write throughput, IOPS, I/O queue depth, utilization percentage, and per-mount filesystem space and inode usage. Predict disk full events with AI capacity forecasting.
Network Interface Monitoring
Per-interface TX/RX bytes, packets, errors, drops, and collisions. Track socket statistics (TCP established, TIME_WAIT, CLOSE_WAIT) for detecting connection leaks and network saturation.
Process-Level Visibility
Top CPU and memory consuming processes, process counts by state (running, sleeping, zombie), and file descriptor usage per process. Detect resource-hungry processes before they impact system stability.
Kernel and System Event Correlation
Ingest dmesg, OOM killer events, and systemd journal alongside host metrics. TigerOps correlates kernel OOM kills with memory metric trends and CPU throttling with cgroup limits.
TigerOps Agent Installation & Config
Install and configure the TigerOps agent on a Linux host in under 2 minutes.
# One-line install (Ubuntu/Debian)
curl -fsSL https://install.atatus.net/agent | sudo bash -s -- \
--api-key ${TIGEROPS_API_KEY} \
--tags "env:production,region:us-east-1,team:platform"
# Or install manually and configure:
# /etc/tigerops-agent/agent.yaml
api_key: ${TIGEROPS_API_KEY}
endpoint: https://ingest.atatus.net
host_tags:
env: production
region: us-east-1
team: platform
# Host metric collection
host_metrics:
enabled: true
collection_interval: 10s
cpu:
per_cpu: true
collect_cpu_time: true
disk:
ignored_fs: [tmpfs, devtmpfs, squashfs]
network:
ignored_interfaces: [lo, docker0]
processes:
top_n: 20 # top 20 processes by CPU/memory
# Log collection
logs:
enabled: true
sources:
- type: file
path: /var/log/syslog
tags: {source: syslog}
- type: file
path: /var/log/auth.log
tags: {source: auth}
- type: journald
units: [] # collect all unitsCommon Questions
Which Linux distributions does the TigerOps agent support?
The TigerOps agent supports Ubuntu 18.04+, Debian 10+, CentOS 7+, RHEL 7+, Amazon Linux 2, Fedora 36+, and Alpine Linux 3.14+. Both x86_64 and arm64 architectures are supported.
How much system overhead does the TigerOps agent add?
The TigerOps agent uses less than 50 MB of RSS memory and under 0.1% CPU on modern hardware. Metrics are collected every 10 seconds by default and can be adjusted down to 1-second granularity for intensive monitoring.
Can I deploy the TigerOps agent without root privileges?
Most host metrics require no root access. A few metrics (network socket statistics, some /proc files) require either root or specific Linux capabilities. The install script creates a tigerops-agent user with the minimum required capabilities.
How do I monitor hundreds of Linux hosts from one TigerOps workspace?
Install the agent on each host with the same API key. Use host_tags in the agent config to label hosts by environment, region, team, and role. TigerOps groups metrics by these tags for fleet-wide dashboards and aggregations.
Does TigerOps support cgroup v2 metrics on modern Linux?
Yes. The TigerOps agent reads cgroup v2 controllers (cpu, memory, io, pids) from /sys/fs/cgroup. This provides accurate per-container and per-service resource usage on systemd-managed hosts running container workloads.
Deep Linux Host Monitoring in Under 2 Minutes
One install command covers CPU, memory, disk, network, processes, and logs. AI capacity forecasting and anomaly detection included.