Command-line interface

The Google Cloud CLI has a group of commands, gcloud logging, that provide a command-line interface to the Logging API. A summary of the important commands and examples of their use are shown on this page.

For additional information, go to the following sources:

• For detailed documentation on the Logging Google Cloud CLI, read the reference pages for the gcloud logging command group. There might also be new or changed commands in the beta command group: gcloud beta logging.
• For documentation on the Logging API, read Cloud Logging API.

Getting started

1. In the Google Cloud console, activate Cloud Shell.

   Activate Cloud Shell

   At the bottom of the Google Cloud console, a Cloud Shell session starts and displays a command-line prompt. Cloud Shell is a shell environment with the Google Cloud CLI already installed and with values already set for your current project. It can take a few seconds for the session to initialize.

2. Set your default project so you don't have to supply the --project flag with each command:

   gcloud config set project PROJECT_ID

After you've installed and initialized the Google Cloud CLI, you can run gcloud logging commands from the command line in the same way you use other command-line tools.

Permissions

gcloud logging commands are controlled by Identity and Access Management (IAM) permissions.

To use any of the gcloud logging commands, you must have the serviceusage.services.use permission. You must also have the IAM role that corresponds to the log's location, and to your use case:

For additional information on required Logging IAM roles and permissions, go to Access control.

Summary of commands

The following sections provide summaries and examples of the gcloud CLI command-line interface for Logging. However, some command options and details are omitted; the online documentation for the gcloud CLI commands is authoritative.

From the command line, you can add --help to a partial command to get more details. For example:

gcloud logging --help
gcloud logging sinks --help
gcloud logging sinks create --help

In a few cases, important command features in the Beta version of the gcloud CLI are available:

gcloud beta logging metrics create --help

Over time, Beta features might be rolled into the standard release and new features might be added to the Beta release.

Log entries

You can write and read log entries using gcloud.

Writing log entries

Use the gcloud logging write command, corresponding to the API method entries.write.

gcloud logging write LOG_NAME  ...

LOG_NAME can either be the LOG_ID or the full resource name of the log with the LOG_ID URL-encoded.

For simplicity, this command makes several assumptions about the log entry. For example, it always sets the resource type to global.

Examples

Write a log entry to log my-test-log in the current project with a plain-text payload and a severity of ERROR:

gcloud logging write my-test-log "A simple entry" --severity=ERROR

Write a log entry with a structured (JSON) payload:

gcloud logging write my-test-log '{ "message": "My second entry", "weather": "partly cloudy"}' --payload-type=json

Write a log entry to a folder, billing account, or organization. The following command writes a log entry to a folder:

gcloud logging write my-folder-log "A folder log entry"  --folder=FOLDER_ID

To find your log entries, look in the Logs Explorer under the Global resource type.

You can also gcloud logging read or the corresponding API method. See the example in Reading log entries.

Reading log entries

To retrieve log entries, use the gcloud logging read command, corresponding to the API method entries.list:

gcloud logging read FILTER ...

To read log entries in folders, billing accounts, or organizations, append the --folder, --billing-account, or --organization options.

Examples

Read up to 10 log entries in your project's syslog log from Compute Engine instances containing payloads that include the word SyncAddress. The log entries are to be shown in JSON format:

gcloud logging read "resource.type=gce_instance AND logName=projects/PROJECT_ID/logs/syslog AND textPayload:SyncAddress" --limit 10 --format json

Following is an example of one returned log entry:

{
  "insertId": "2024-04-07|08:56:48.137651-07|10.162.32.129|-1509625619",
  "logName": "projects/[PROJECT_ID]/logs/syslog",
  "resource": {
    "labels": {
      "instance_id": "15543007601548829999",
      "zone": "global"
    },
    "type": "gce_instance"
  },
  "textPayload": "Apr  7 15:56:47 my-gce-instance google-address-manager: ERROR SyncAddresses exception: HTTP Error 503: Service Unavailable",
  "timestamp": "2024-04-07T15:56:47.000Z"
}

To list logs in a folder, add the --folder flag. That flag restricts the log entries read to only those in the folder itself. The same applies to the --organization and --billing-account flags. The following command retrieves activity-type audit logs from a folder. Log entries are to be shown in the default YAML format:

gcloud logging read "resource.type=folder AND logName:cloudaudit.googleapis.com%2Factivity" --folder=FOLDER_ID

Following is an excerpt of one returned entry. It records a call to SetIamPolicy on the folder:

insertId: mhcr1tc16u
logName: folders/FOLDER_ID/logs/cloudaudit.googleapis.com%2Factivity
protoPayload:
  '@type': type.googleapis.com/google.cloud.audit.AuditLog
  methodName: SetIamPolicy
  ...
  serviceName: cloudresourcemanager.googleapis.com
  status: {}
resource:
  labels:
    folder_id: 'FOLDER_ID'
  type: folder
severity: NOTICE
timestamp: '2024-03-19T16:26:49.308Z'

The previous section, Writing log entries, contains an example of writing a log entry to a folder. Following is the command to read the log entry:

gcloud logging read "resource.type=global" --folder=FOLDER_ID --limit=1

Here is the result:

insertId: 1f22es3frcguaj
logName: folders/FOLDER_ID/logs/my-folder-log
receiveTimestamp: '2024-03-19T18:20:19.306598482Z'
resource:
  type: global
textPayload: A folder log entry
timestamp: '2024-03-19T18:20:19.306598482Z'

Logs

A log, or log stream, is the set of log entries that have the same full resource name. The full resource name is equivalent to the LogName field in the LogEntry.

The full resource name for a log must be one of the following:

projects/PROJECT_ID/logs/LOG_ID
organizations/ORGANIZATION_ID/logs/LOG_ID
folders/FOLDER_ID/logs/LOG_ID
billingAccounts/BILLING_ACCOUNT_ID/logs/LOG_ID

LOG_ID can only contain alphanumeric characters, forward-slash, underscore, hyphen, and period characters. For example, compute.googleapis.com/activity_log is a valid LOG_ID. The full resource name includes the project, folder, billing account, or organization in which the log is located. For more information, go to Google Cloud resource hierarchy.

To manage logs, use the gcloud logging logs command group:

gcloud logging logs list  ...
gcloud logging logs delete  ...

URL Encoding

When passing the full resource name of a log to a gcloud logging command, you must URL-encode the LOG_ID. As an example, a LOG_ID of compute.googleapis.com/activity_log encodes to compute.googleapis.com%2Factivity_log.

When passing the LOG_ID to a gcloud logging command, you don't need to URL-encode the LOG_ID. In this case, the gcloud command encodes for you.

When receiving data that includes the full resource name of a log from a gcloud logging command, the LOG_ID is URL-encoded.

Creating logs

You create a log by writing a log entry to it. See Writing log entries.

Listing log names

Use the gcloud logging logs list command. It executes the API method projects.logs/list.

Only logs that contain log entries are displayed. The display lists full resource names of logs with the LOG_ID URL-encoded.

You can only use gcloud logging logs list to list logs in projects; you can't use it to list logs in folders, billing accounts, or organizations.

Example

List the logs in the current project:

gcloud logging logs list

Sample result:

NAME
projects/my-gcp-project-id/logs/apache-error
projects/my-gcp-project-id/logs/cloudaudit.googleapis.com%2Factivity
projects/my-gcp-project-id/logs/cloudaudit.googleapis.com%2Fdata_access
projects/my-gcp-project-id/logs/compute.googleapis.com%2Factivity_log
projects/my-gcp-project-id/logs/compute.googleapis.com%2Fsystem_log
projects/my-gcp-project-id/logs/syslog

Deleting logs

To delete logs, use the gcloud logging logs delete command. It decides, based on the full resource name of a log, which of the following delete API commands to execute: projects.logs, organizations.logs, folders.logs, or billingAccounts.logs.

Example: Delete a project log

To delete a log that exists in the current project using the LOG_ID:

gcloud logging logs delete my-new-log

Really delete all log entries from
[my-new-log]?g_logs
Do you want to continue (Y/n)?  Y
Deleted [my-new-log].

An alternative is to specify the full resource name of a log with the LOG_ID URL-encoded:

gcloud logging logs delete projects/my-gcp-project-id/logs/compute.googleapis.com%2Factivity_log

Really delete all log entries from
[projects/my-gcp-project-id/logs/compute.googleapis.com%2Factivity_log]?
Do you want to continue (Y/n)?  Y
Deleted [projects/pamstestproject1/logs/compute.googleapis.com%2Factivity_log].

Example: Delete logs in folders, billing accounts or organizations

To delete a log contained in folders, billing accounts or organizations, pass the full resource name with the LOG_ID URL-encoded. For example, the following command deletes a log in a folder:

gcloud logging logs delete folders/FOLDER_ID/logs/my-folder-log

FOLDER_ID must be the unique number that identifies the folder.

Live tailing log entries

Live tailing lets you view your log entries in real time as Cloud Logging writes them, by using either the Google Cloud CLI or the Cloud Logging API.

Live tailing isn't supported for log buckets with field-level access controls, however you can stream logs for those buckets in the Logs Explorer.

For information on the API method for live tailing, see the entries.tail method.

Installing gcloud alpha logging tail

To use gcloud alpha logging tail, you need to have Python 3 and the grpcio Python package installed.

For instructions on how to install Python, see the Python page. For instructions on how to install the Python package manager, pip, that is needed to install the grpcio package, see The Python Package Installer page.

Complete the following steps to install gcloud alpha logging tail:

1. Verify that you have the Google Cloud CLI installed. For instructions on how to install the Google Cloud CLI, see Installing Google Cloud CLI.

2. Verify that you're using version 302.0.0 or greater of the gcloud CLI.

   gcloud version
   

   For instructions on updating the gcloud CLI, see gcloud components update.

3. Install the gcloud CLI alpha components:

   gcloud components install alpha
   

4. For MacOS, Linux, and Cloud Shell users:

   1. Install gRPC client libraries:

       sudo pip3 install grpcio
      

   2. Set the environment variable CLOUDSDK_PYTHON_SITEPACKAGES to any value:

       export CLOUDSDK_PYTHON_SITEPACKAGES=1
      

5. Use the following commands to set your Google Cloud project ID and to authenticate:

   gcloud config set project PROJECT_ID
   gcloud auth login
   

   To get the project ID, see Creating and managing projects.

6. Verify that gcloud alpha logging tail is installed by running the following command:

   gcloud alpha logging tail
   

   The command displays the following message:

   Initializing tail session.

   You are now viewing the log entries for your Google Cloud project as Logging writes them.

   

For more information on using live tailing, see the gcloud alpha logging tail reference guide.

Buffering and ordering

Because Logging can receive log entries out of chronological order, live tailing provides a buffer-window setting so you can balance the tradeoff between viewing the log entries as they are being written and viewing them in ascending order. You can set the buffer window between 0 and 60 seconds.

Note the following characteristics of the buffer window:

• The default buffer window is two seconds.

• Logging delays writing the log entries to log buckets for the duration of the buffer window.

• If a log entry is written outside of the buffer window, then Logging returns the log entries as they are received.

When configuring the buffer window, you make a tradeoff between viewing logs as they are written and viewing the entries out of order.

Limits and quotas

The following table lists the limits and quotas for live tailing:

Client limitations

For a Google Cloud project that writes lots of entries quickly, your client might be unable to consume them as quickly as they're being written. In this case, Logging limits the total number of entries sent, prioritizing the most recent entries. At the end of the tail session, Logging returns the number of entries that were not displayed due to the limits of the client.

Resource descriptors

All log entries contain an instance of one of a fixed set of monitored resource types that generally identifies the resource the log entry comes from, such as a particular Compute Engine VM instance. For a list of monitored resource types, go to Monitored Resource List.

To list the current resource descriptor types, use the gcloud logging resource-descriptors list command, corresponding to the API method monitoredResourceDescriptors.list. You don't need any special permissions to list the resource types.

Examples

List all the resource types that have instance in their names:

gcloud logging resource-descriptors list --filter="type:instance"

Result:

TYPE                        DESCRIPTION                                                 KEY
gce_instance                A virtual machine instance hosted in Compute Engine. project_id,instance_id,zone
spanner_instance            A Cloud Spanner instance.                                   project_id,instance_id,location,instance_config
redis_instance              A Redis instance hosted on Google Cloud MemoryStore.        project_id,region,instance_id,node_id
gce_instance_group          A Compute Engine instance group resource.            project_id,instance_group_id,instance_group_name,location
gce_instance_group_manager  A Compute Engine instance group manager resource.    project_id