Logs & Metrics

Overview

Logs in Tapitalee provide real-time and historical access to your application’s output, error messages, and system events. The logging system captures output from all your application processes, tasks, and system components, making it easy to monitor application behavior, debug issues, and track system performance.

Log Sources

Application Logs

• Process Output: Standard output (stdout) and error (stderr) from your application processes

System Logs

• Container Lifecycle: Container start, stop, and restart events. Deployment logs.

Commands

Show Application Logs

Show CloudWatch logs for a process or task

tapit show logs [task=task_name] [process=process_name] [container=primary] [limit=lines] [filter=string] [from=iso8601] [to=iso8601] [interval=1h|4h|12h|24h|3d|7d|14d] [-f|--follow]

Parameters

• task: Show logs for a specific task instead of the main process
• process: Show logs for a specific named process (default: default process)
• container: Show logs for a specific container name (default: primary)
• limit: Maximum number of log lines to return (default: 50, max: 10000, not with -f)
• filter: Only show log lines containing this string (past logs only, not with -f)
• from: Start of the time range (ISO8601 timestamp)
• to: End of the time range (ISO8601 timestamp, defaults to now)
• interval: Named interval instead of from/to (eg. 4h, 24h, 7d)
• -f|--follow: Continuously tail and stream new log entries

Examples

# View recent application logs
tapit show logs

# View logs for specific process
tapit show logs process=default

# Follow logs in real-time
tapit show logs process=worker --follow

# View logs for specific task
tapit show logs task=task-123

# View logs for specific container
tapit show logs process=web container=sidecar

# Only show lines containing both words
tapit show logs filter='timeout retry'

Filtering Logs

The filter parameter uses the AWS CloudWatch Logs filter pattern syntax.

Filter	Meaning
filter='word1 word2'	Lines containing both words
filter='"word1 word2"'	Lines containing the exact phrase, including the space
filter='?word1 ?word2'	Lines containing either word (optional terms)
filter='included -excluded'	Lines containing included but not excluded
filter='%Error.*format%'	Regular expression match, delimited by %

Datadog Log Redirection

If a [Datadog add-on]({{< relref “/add-ons/datadog” >}}) with enable_logging=true is attached, your primary container logs are shipped to Datadog instead of CloudWatch. When the add-on also has a datadog_app_key set, show logs and the web Logs tab transparently fetch those logs from the Datadog API — including live tail and from/to time ranges. In that case the filter parameter uses Datadog log search syntax rather than the CloudWatch pattern syntax above.

Log Retention

• Logs are kept for 90 days in Cloudwatch Logs

Metrics

CloudWatch metrics for a process or add-on can be shown as a table over a chosen time range and interval.

Show Metrics

Show CloudWatch metrics for an add-on or process as a table

tapit show metrics addon=addon_name|process=process_name [from=iso8601] [to=iso8601] [interval=1h|4h|12h|24h|3d|7d|14d]

Parameters

• addon: Name of the add-on to show metrics for
• process: Name of the process to show metrics for
• from: Start of the time range (ISO8601 timestamp)
• to: End of the time range (ISO8601 timestamp, defaults to now)
• interval: Named interval instead of from/to (eg. 4h, 24h, 7d)

Examples

# Metrics for a process
tapit show metrics process=default

# Metrics for an add-on
tapit show metrics addon=my-database

# Metrics over a preset time range
tapit show metrics process=worker interval=24h

# Metrics over an explicit time range
tapit show metrics addon=my-database from=2026-01-01T00:00:00Z to=2026-01-02T00:00:00Z