isrv

Simple anonymous and temporary file sharing service.

Visit https://isrv.nl to see it in action.

Description

isrv is a lightweight file sharing service that provides anonymous temporary storage with customizable expiration times. Users can upload files and share them via generated links without requiring registration or personal information.

Goals

• Anonymous temporary storage, with customizable expiration time
• Easy installation: Single statically-linked binary that can be deployed anywhere
• Manage your own uploads, without compromising anonymity

Roadmap

This project is a work in progress, here's a list of things I am working on in no particular order:

• True E2EE with the key in the URL fragment; The server never sees the key and cannot decrypt files
• Client mode in the same binary, allowing users to upload and manage their files with ease
• Metadata stripping for uploaded files (EXIF, ID3, etc.)
• Support non-temporary storage option for users who want to keep their files indefinitely
• Built-in TLS support with automatic certificate management CertMagic
• "Burn After Reading" feature: Expiry after N downloads
• Zip file support for multiple uploads:
  • Upload multiple files and download them as a single zip file
• Text/Paste file preview in the web interface with syntax highlighting
• Improvements to abuse detection and mitigation
• Storage tiers (namely "hot" and "cold") to optimize costs and performance based on file access patterns

Installation

Pre-built binaries

Download the latest release for your platform from the releases page and make it executable:

# Linux
wget https://github.com/markhc/isrv/releases/latest/download/isrv-linux-amd64
chmod +x isrv-linux-amd64
sudo mv isrv-linux-amd64 /usr/local/bin/isrv

To run isrv as a system service (auto-start on boot, restart on failure), the binary can install itself as a systemd unit:

sudo isrv install

See docs/systemd.md for details, upgrades, and removal.

Docker

• Create a docker-compose.yaml file (you can use the project's docker-compose.yaml as base)
• Create the data folder where the database and files will be kept.
• Start the container

Example:

mkdir -p ./data
docker compose up -d

From source

Requires Go 1.25 or later:

git clone https://github.com/markhc/isrv.git
cd isrv
make build

The binary will be available in the build/ directory.

Usage

Running the server is as easy as starting the binary.

# Generates a default configuration file on $HOME/.config/isrv/config.yaml
isrv makeconf

# Starts the webserver (will load config file if it exists)
isrv

# Starts the webserver with a specific config file
isrv serve -c config.yaml

# Installs and starts isrv as a systemd service (see docs/systemd.md)
sudo isrv install

If no config file is provided the application will look for one in standard places and, if none can be found, default values will be used.

The web interface will be available at http://localhost:8080.

Configuration

Configuration can be provided via:

• Configuration file
• Environment variables

Configuration file

default_config.yaml

Environment Variables

When set, environment variables override the corresponding values from the configuration file.

Variable	Default	Description
ISRV_SERVER_NAME	iSRV	Sets the server name
ISRV_SERVER_URL	http://localhost:8080	Sets the server URL
ISRV_SERVER_HOST	0.0.0.0	Sets the server host address
ISRV_SERVER_PORT	8080	Sets the server port
ISRV_STORAGE_TYPE	local	Storage backend: local, s3 or gcs
ISRV_STORAGE_PATH	-	Sets the storage base path (directory for local, key prefix for s3/gcs)
ISRV_STORAGE_BUCKET_NAME	-	Bucket name for s3 and gcs storage
ISRV_STORAGE_CREDENTIALS_FILE	-	Service account JSON key file for gcs storage (empty = Application Default Credentials)
ISRV_LOGGING_LOG_TO_FILE	false	Whether we should log to a file
ISRV_LOGGING_LOG_IPS	true	Log uploaders IP
ISRV_LOGGING_ANONYMIZE	false	Minimal-logging mode: drop successful request logs and omit all identifying fields (supersedes logIps)
ISRV_LOGGING_PATH	-	Sets the log file path
ISRV_RANDOM_ID_LENGTH	12	Sets the length of randomly generated file IDs
ISRV_MAX_FILE_SIZE_MB	512	Sets the maximum file size in megabytes
ISRV_CLEANUP_ENABLED	true	Enable the job that removes expired files
ISRV_CLEANUP_INTERVAL	1m	The interval at which the cleanup job runs
ISRV_ADMIN_USERNAME	-	Admin panel username (enables the panel when set together with the password)
ISRV_ADMIN_PASSWORD	-	Admin panel password
ISRV_ADMIN_SESSION_SECRET	-	HMAC key for signing admin session cookies (random if unset)
ISRV_ADMIN_FAILED_LOGIN_LIMIT	5	Max failed admin login attempts per hour before the IP is blocked
ISRV_ADMIN_FAILED_LOGIN_BLOCK_DURATION	12h	How long an IP is blocked after exceeding the failed login limit

Reverse proxy

isrv does not terminate TLS, so production deployments normally run behind a reverse proxy. See docs/reverse-proxies/nginx.md for a complete NGINX server block and guidance on client IP detection, upload/download buffering, and which endpoints to keep private.

Anonymity-focused deployments

For running isrv as a Tor onion service, along with the no-logs anonymize logging mode and the privacy trade-offs involved, see docs/tor.md.

Multi-replica / high availability

isrv defaults to a single replica with zero external dependencies. To run several replicas behind a load balancer (e.g Kubernetes), see docs/multi-replica.md.

Observability endpoints

When the server is running, the following infrastructure endpoints are always available:

Endpoint	Description
GET /healthz	Liveness probe; always returns 200 {"status":"ok"}
GET /readyz	Readiness probe; returns 200 when the database and storage backend are both reachable, otherwise 503 with a per-check error map
GET /metrics	Prometheus scrape endpoint

Admin Panel

A single-administrator panel lets you view, search, preview and delete uploaded files. It is reachable at /admin

The panel is disabled unless both an admin username and password are configured.

admin:
  username: "admin"
  password: "change-me"
  # Optional: HMAC key for signing session cookies. If omitted, a random key is
  # generated at startup and existing sessions are invalidated on every restart.
  sessionSecret: ""
  # Optional: how long a login session stays valid (default 24h).
  sessionTtl: 24h
  # Max failed login attempts per hour before the offending IP is blocked.
  failedLoginLimit: 5
  # How long to block IPs that exceed the failed login limit (backs off
  # exponentially for repeated offenses).
  failedLoginBlockDuration: 12h

Development

Building

Build for current platform:

make build

Testing

Run tests:

make test

Run tests with coverage:

make test-coverage

Development workflow

Spin up a development server with hot reload:

make dev

AI Usage

I believe AI disclosure is important, and as such I should say that this project does make use of AI tools.

Test cases, documentation and some code snippets have been generated with the help of tools such as Claude Code and GitHub Copilot. Furthermore, code reviews are currently performed by Coderabbit.

Contributions from the community are welcome. However, please disclose any AI usage in your pull requests. Contributions deemed to be AI generated might be rejected if they do not meet the quality standards of the project.