canister_logs

Canister logs

🥷 Try it live — no local setup. ICP Ninja is a web-based IDE that builds and deploys this project to the mainnet for free, right in your browser. Click the badge above, or hit Deploy if you're already in Ninja. To build and run it locally instead, follow the steps below.

This example demonstrates canister logging and error handling on the Internet Computer. It shows how to use ic_cdk::println! for diagnostic output, how traps, panics, and memory errors are automatically captured in the canister log, and how periodic timers generate log entries.

The canister exposes methods to print messages, trigger explicit traps, cause panics, simulate stable-memory out-of-bounds errors, and call raw_rand. A key behavioral distinction is demonstrated: log messages from update calls and replicated queries are recorded in the canister log, while messages from non-replicated query calls are not.

Build and deploy from the command line

Prerequisites

• Node.js
• icp-cli: npm install -g @icp-sdk/icp-cli @icp-sdk/ic-wasm

Install

git clone https://github.com/dfinity/examples
cd examples/rust/canister_logs

Deploy and test

icp network start -d
icp deploy
bash test.sh
icp network stop

Check canister logs

After deploying, inspect logs directly:

icp canister logs backend

Expect to see periodic timer trap entries every 5 seconds:

[0. 2024-05-23T08:32:26Z]: right before timer trap
[1. 2024-05-23T08:32:26Z]: [TRAP]: timer trap
[2. 2024-05-23T08:32:31Z]: right before timer trap
[3. 2024-05-23T08:32:31Z]: [TRAP]: timer trap

To follow logs in real time:

icp canister logs --follow backend

Call canister methods manually

# Print a message (update call — recorded in logs)
icp canister call backend print '("hello!")'

# Print via replicated query (recorded in logs)
icp canister call backend print_query '("hello from query!")'

# Print via non-replicated query (NOT recorded in logs)
icp canister call --query backend print_query '("this will not appear in logs")'

# Trigger an explicit trap (error returned, message recorded in logs)
icp canister call backend trap '("oops!")'

# Trigger a Rust panic (error returned, message recorded in logs)
icp canister call backend panic '("something went wrong")'

# Trigger a stable-memory out-of-bounds error (recorded in logs)
icp canister call backend memory_oob '()'

# Trigger a failed unwrap (recorded in logs)
icp canister call backend failed_unwrap '()'

# Call raw_rand (result returned, success message recorded in logs)
icp canister call backend raw_rand '()'

Security considerations and best practices

When building production canisters, review the ICP security best practices.