Backed by mandō

Real-time log streaming.
Zero blind spots.

Tailstream is the observability platform that outputs the last N lines of any file and follows it as it grows. Real-time streaming. Process-aware lifecycle. Auto-reconnect on rotation. This is production-grade log intelligence.

Watch the demo
S K D R T
9,400+ SREs streaming production logs
~/production
$ tailstream -f /var/log/app.log
[INFO] Server started on :8080
[INFO] Connected to database cluster
[INFO] Health check passed
[INFO] GET /api/users 200 12ms
[INFO] POST /api/orders 201 45ms
[WARN] Slow query detected: 892ms
[INFO] GET /api/products 200 8ms
[INFO] WebSocket connection established
[ERROR] Connection pool exhausted, retrying...
[INFO] Connection pool recovered
[INFO] GET /api/health 200 2ms
Live Stream

Trusted by observability teams at

🟣 Datadog
🟠 Grafana
🔵 Kubernetes
Vercel
🔷 Docker
🟢 Splunk
🐙 GitHub Actions
🔶 PagerDuty

Every flag. Every stream mode.
Production hardened.

We took the core primitives of log tailing and turned them into an observability platform that never misses a line.

📡

LiveFollow™ -f

Real-time log streaming that outputs appended data as the file grows. Attach to any log file and watch events arrive the instant they're written. This is the core of production observability.

🔄

ResilientStream™ -F

Follow by name with automatic retry. When log files rotate, get renamed, or get recreated, ResilientStream™ detects the change and re-attaches automatically. Zero downtime. Zero missed lines.

📊

LastLines™ -n NUM

Output the last NUM lines of any file. Default 10. Want the last 500 lines of your error log? Done. The precision tail that gives you exactly the data window you need.

🔢

ByteSlice™ -c NUM

Output the last NUM bytes instead of lines. Byte-level precision for binary-aware workflows. Use +NUM to start from a specific byte offset. Surgical data extraction.

💀

ProcessGuard™ --pid=PID

Attach your stream to a process lifecycle. Tailstream automatically terminates when the tracked PID dies. Perfect for CI jobs, build processes, and ephemeral workloads.

⏱️

PollTune™ -s SECS

Configure the sleep interval between follow iterations. Fine-tune your polling frequency from millisecond-fast to battery-saving slow. Your stream, your cadence.

🔁

RetryConnect™ --retry

Keep trying to open a file even if it's temporarily inaccessible. Combined with LiveFollow™, this creates an indestructible stream that survives log rotation, disk pressure, and file system hiccups.

OffsetStart™ +NUM

Use -n +NUM to start output from line NUM, or -c +NUM to start from byte NUM. Skip headers, jump to the action, and begin streaming from any point in the file.

How NeonScale achieved 99.99% observability coverage with Tailstream

"We were running 340 microservices across 12 Kubernetes clusters. Our previous log aggregation solution had a 4-second ingestion delay and dropped lines during log rotation. We needed real-time."

NeonScale's SRE team deployed Tailstream's ResilientStream™ across every pod. With -F handling log rotation and --pid tracking container lifecycles, they eliminated 100% of their log gaps.

Result: MTTR dropped from 14 minutes to 90 seconds. Zero missed log lines across 2.4 billion events per day.

90s
MTTR (down from 14min)
0
Dropped log lines
340
Services monitored
P
Priya Ramaswamy VP of Platform Engineering, NeonScale

From file to stream
in milliseconds.

01

Point at your log files

Pass one or more file paths to Tailstream. Supports local files, rotated logs, and stdin. If no file is specified, Tailstream reads from standard input.

02

Configure your stream window

Set your tail depth with -n (lines) or -c (bytes). Default is the last 10 lines. Use +NUM prefix to start from a specific offset instead of the end.

03

Enable real-time follow

Add -f to stream appended data as it arrives, or -F for resilient streaming that survives log rotation. Attach --pid to bind the stream to a process lifecycle.

04

Stream to stdout in real time

Log events flow to your terminal the instant they're written. Pipe into grep, awk, or your alerting pipeline. Zero buffering delay. Production-grade throughput.

0
Log lines streamed daily
0
Stream uptime SLA
0
Average stream latency
0
Production streams active now

Start free. Scale to production.

Starter
$0/mo

Last 10 lines. The basics.

  • Default 10-line tail
  • Single file streaming
  • Quiet mode (-q)
  • Verbose headers (-v)
  • Community support
Enterprise
$299/mo

Resilient streaming + process tracking

  • Everything in Pro
  • ResilientStream™ (-F)
  • ProcessGuard™ (--pid)
  • RetryConnect™ (--retry)
  • Multi-file tailing
  • SSO & RBAC
  • Dedicated CSM
On-Prem
Custom

Self-hosted log intelligence

  • Everything in Enterprise
  • Air-gapped deployment
  • Custom SLAs
  • Compliance certs
  • White-glove onboarding

SREs who stream
never go back.

★★★★★

"We replaced our entire log aggregation sidecar with Tailstream's LiveFollow™. Real-time event streaming with zero ingestion delay. Our on-call engineers can now see errors the moment they happen, not 4 seconds later."

J
Jordan Whitfield Staff SRE, ScaleForge
★★★★★

"ResilientStream™ solved log rotation for us permanently. We rotate logs every 15 minutes across 200 nodes. Before Tailstream, we'd lose 2-3 seconds of events on every rotation. Now? Zero gaps. Zero."

A
Amara Okonkwo VP Engineering, DataPipe
★★★★★

"ProcessGuard™ with --pid is a game-changer for CI/CD observability. We attach Tailstream to every build process and it automatically terminates when the job finishes. Clean lifecycle management."

M
Marcus Chen Principal Engineer, NeonCloud
★★★★★

"The -n +1 feature is criminally underrated. Skip the header line and stream the rest of the file in real time. We use it to process CSV exports as they're being generated. Tailstream is infrastructure."

R
Cole Dupont DevOps Lead, InfraCorp
★★★★★

"We evaluated Splunk, Datadog, and Tailstream for real-time log tailing. Tailstream was the only one that delivered sub-millisecond latency with zero configuration. Pipe it into grep and you have a monitoring stack."

T
Tyler Nakamura CTO, StreamKit

See Tailstream in action.

Real output. Real log streaming. Zero smoke and mirrors.

~/production

Questions from
on-call engineers.

LiveFollow™ (-f) follows a file descriptor — even if the file is renamed, it continues tracking the original file. ResilientStream™ (-F) follows by name and retries — if the file is rotated, deleted, or recreated, it detects the change and re-attaches. Use -F for production log files that rotate. Use -f for files that stay in place.

When you pass --pid=PID along with -f, Tailstream monitors the specified process. When that process terminates, Tailstream automatically exits. This is invaluable for CI/CD pipelines, build jobs, and any ephemeral workload where you want the stream to die with the process.

Yes. Use -n +NUM to start output from line NUM, or -c +NUM to start from byte NUM. For example, -n +1 outputs the entire file, while -n +100 skips the first 99 lines and streams everything from line 100 onward.

Absolutely. Pass multiple file paths and Tailstream will output the last lines of each, preceded by a header with the file name. Combined with -f, you can follow multiple log files simultaneously. Use -q to suppress headers or -v to always show them.

With --retry, Tailstream will keep attempting to open the file. Combined with -F, this creates an indestructible stream that waits for the file to appear, attaches when it's created, and re-attaches if it's rotated. Ideal for log files that haven't been created yet at service startup.

Real-time log intelligence
starts here.

Join 9,400+ SREs and platform engineers who've already secured their place on the Tailstream platform. Early access members get lifetime Pro pricing.

No spam. Just real-time observability updates.

See mandō's portfolio

The accelerator behind the tools that run the internet.