Serverless Computing usage and billing

Permissions

To estimate Serverless Computing resource consumption and costs from the slow query log, you need specific permissions. The following sections detail the required permissions and authorization methods.

• Query the data scan volume for all databases in an instance

  • Option 1: Grant the Superuser permission. The following command grants this permission:

    -- Replace "Alibaba Cloud account ID" with the actual username. For a RAM user, prepend the account ID with "p4_".
    ALTER USER "Alibaba Cloud account ID" SUPERUSER;

  • Option 2: Add the user to the pg_read_all_stats user group. Use one of the following commands based on your permission model:

    Note

    In addition to the Superuser permission, Hologres allows members of the pg_read_all_stats user group to view the data scan volume for all databases. If a regular user needs to view all logs, they can ask a Superuser to grant them this permission and add them to the group.

    GRANT pg_read_all_stats TO "Alibaba Cloud account ID"; -- For the expert permission model
    CALL spm_grant('pg_read_all_stats', 'Alibaba Cloud account ID');  -- For the simple permission model (SPM)
    CALL slpm_grant('pg_read_all_stats', 'Alibaba Cloud account ID'); -- For the schema-level simple permission model (SLPM)

• View the data scan volume for the current database

  • Enable the simple permission model (SPM) or the schema-level simple permission model (SLPM) and add the user to the db_admin role. The db_admin role can view the data scan volume for the current database. Use one of the following commands to grant this role.

    CALL spm_grant('<db_name>_admin', 'Alibaba Cloud account ID');  -- For SPM
    CALL slpm_grant('<db_name>.admin', 'Alibaba Cloud account ID'); -- For SLPM

  • A regular user can query the data scan volume for queries that they have executed in the current database under their own account.

Query consumption for a single SQL

Use the following SQL statement to query the detailed resource consumption of a Serverless Computing job.

SELECT
    *,
    queue_time_ms,	-- The time a SQL query spends waiting in the Serverless Computing queue, in milliseconds (ms).
    serverless_allocated_cores,	-- The number of CUs actually allocated to the current SQL query by Serverless Computing.
    serverless_resource_used_time_ms, -- The duration that the SQL query actually occupies Serverless Computing resources, in milliseconds (ms).
    (serverless_allocated_cores::DECIMAL(38, 4)) * (serverless_resource_used_time_ms::DECIMAL(38, 4)) AS serverless_cums
FROM
    hologres.hg_serverless_computing_query_log;

Query consumption over a time range

Use the following SQL statement to query the total resource consumption of successful Serverless Computing jobs within a specified time range.

SELECT
    (SUM((serverless_allocated_cores::DECIMAL(38, 4)) * (serverless_resource_used_time_ms::DECIMAL(38, 4))) / 1000 / 60 / 60)::bigint AS serverless_cuh
FROM
    hologres.hg_serverless_computing_query_log
WHERE
    status = 'SUCCESS'
    AND serverless_allocated_cores IS NOT NULL
    AND serverless_resource_used_time_ms IS NOT NULL
    AND query_end BETWEEN '2024-05-01 00:00:00+08'::timestamptz AND '2024-05-31 24:00:00+08'::timestamptz;-- Specify the time range.

Set a daily usage limit

Starting with Hologres V3.1.5, you can use SQL to set a daily limit on Serverless Computing resource usage. Only a Superuser can set this limit.

-- Set a limit for a database.
ALTER DATABASE <db_name> SET hg_serverless_computing_daily_max_cuh_usage_threshold = <value>;
-- Set a limit for a user.
ALTER USER <user_name> SET hg_serverless_computing_daily_max_cuh_usage_threshold = <value>;

• Limit details

  • The default value is -1, which means no limit is set.

  • The unit is CU·H.

  • Hologres aggregates and updates the total Serverless Computing resource usage for the current instance every 10 minutes. Therefore, there may be a delay of up to 10 minutes in enforcing the limit.

  • Before executing a SQL query, Hologres compares the total serverless resources used today by the instance with the serverless resource limit for the current user or database. If the limit is reached, the query automatically falls back to using the instance's compute resources. If the limit has not been reached, the query runs by using serverless resources as normal.

  • Hologres calculates the daily usage of Serverless Computing resources based on the natural day in the default time zone of the current database. You can run the following commands in the same session to query the default time zone of the current database.

    -- Reset the time zone to avoid influence from the client or user's default time zone.
    RESET timezone;
    -- Query the default time zone of the current database.
    SHOW timezone;

• Recommendations

  • If you are a long-term user of the Hologres Serverless Computing feature, you can determine an appropriate daily usage limit by querying your resource consumption over a specific time range.

  • If you have not used the Hologres Serverless Computing feature before, we recommend that you first use the feature until your workload stabilizes. Then, based on your usage patterns, decide whether to set a daily usage limit and determine an appropriate value.

If you do not want SQL queries to fall back to instance compute resources when the Serverless Computing limit is reached, you can run the following command to make them fail immediately. This parameter can be set only by a Superuser.

-- Set the parameter at the database level.
ALTER DATABASE <db_name> SET hg_serverless_computing_enable_fallback_when_exceed_cuh_threshold = false;
-- Set the parameter at the user level.
ALTER USER <user_name> SET hg_serverless_computing_enable_fallback_when_exceed_cuh_threshold = false;

Cost analysis

You can use the Cost > Cost Analysis module in Expenses and Costs to monitor and analyze the costs for Hologres Serverless Computing compute resources.

1. Log on to Expenses and Costs. In the left-side navigation pane, click Cost > Cost Analysis.

2. On the Cost Analysis page, in the Condition List > Classification Dimension section on the right, select Billing Item. Then, select a Cost Type and Time Granularity based on your needs.

3. In the Filter Conditions area on the right side of the Cost Analysis page, set Billing Item to Serverless Computing compute resource, Hologres Dedicated Instance (Pay-As-You-Go) , and click Apply Filter Conditions. You can then view the Serverless Computing costs for the specified time range.

For more information, see Cost Analysis.

Spending alerts

You can use Alibaba Cloud Expenses and Costs to manage budgets and set alerts for pay-as-you-go spending on specific products in certain regions. For more information, see Expense Alerts.

Alerts for single SQL duration

Hologres supports runtime metrics for Serverless Computing. We recommend creating alert rules based on your business use cases to avoid unexpected costs.

For example, for the metric Longest duration among running Serverless Computing queries, we recommend the following alert rule:

Warn: "The 'Longest duration among running Serverless Computing queries' is greater than or equal to 3,600,000 milliseconds for 5 consecutive periods (1 period = 1 minute)."

For more information, see Monitoring and alerting best practices.