This topic describes how to use the application management feature to create applications (job groups), test worker connectivity, and grant permissions to RAM users.
Create an application
You must create an application before creating a scheduling job.
Step 1: Basic configuration
-
General application: Select this option for applications that are not deployed on Kubernetes or that do not require Kubernetes jobs.
-
Kubernetes application: Select this option for applications that are deployed on Kubernetes and require Kubernetes jobs.
-
This parameter is not available for the Basic Edition of general applications.
-
If no write operations are performed on the log service for more than one year, the service is automatically disabled. You must re-enable it manually.
-
Financial day: A trading day for financial services.
-
Workday: A standard day of work or business operations.
-
Log in to the MSE SchedulerX console.
-
In the top navigation bar, select a region.
-
In the left-side navigation pane, click Application management.
-
In the left-side navigation pane, choose Task scheduling, and then click Application management.
-
On the Application management page, select the target namespace, and then click Create Application.
-
In the Basic configuration step, enter the Application name and Application ID, select an Application type, configure advanced settings if needed, and then click Next.
|
Category |
Parameter |
Description |
Default |
|
Basic information |
Application name |
Specifies a custom name for the application. |
None |
|
Application ID |
The application ID is the GroupID used for application access. This ID must be unique within the same namespace; otherwise, the creation fails. You can use the same value as the Application name. |
None |
|
|
Description |
Specifies a description for the application. |
None |
|
|
Application type |
|
General application |
|
|
Version |
Select an edition based on your requirements. For more information, see Billing for MSE SchedulerX instances. |
Professional Edition |
|
|
Simple Log Service |
After enabling this feature and configuring Log4j or Logback, you can view business logs from each job run in the console to troubleshoot issues. Note
|
Enabled |
|
|
Worker busy configuration |
load5 |
The system load average over the past 5 minutes. If the load exceeds this value, the worker is considered busy. A value of 0 sets the threshold to the number of CPU cores on the worker. |
0 |
|
Memory usage |
If the average memory usage of the process over the past 5 minutes exceeds this threshold, the worker is considered busy. |
90% |
|
|
Disk usage |
If the disk usage exceeds this threshold, the worker is considered unhealthy and its status is set to Busy. |
95% |
|
|
Whether to trigger a busy machine |
Specifies whether to dispatch jobs to the worker when it is busy. |
Enabled |
|
|
Advanced configuration |
Maximum number of jobs |
The maximum number of jobs allowed in a job group. |
1000 |
|
Auto scaling |
Specifies whether to enable auto scaling. If enabled, you must also set Global job count. |
Disabled |
|
|
Flow control |
Specifies whether to enable flow control. If enabled, you must also set Concurrent job instances. |
Disabled |
|
|
Concurrent job instances |
The maximum number of job instances that can run concurrently for this application. Job instances that exceed this limit are queued for later execution instead of being discarded. |
0 |
|
|
Calendar |
From the drop-down list, you can schedule the application on a Financial day or a Workday. |
0 |
Step 2: Notification configuration
-
In the Notification configuration step, configure the settings, select a Notification channel, and set the Contacts.
-
Supported notification channels include SMS, Email, Webhook, and Phone.
For more information about configuring a webhook, see Webhook configuration.
-
You can notify a Contact group or an individual Contact.
-
Contact group
All contacts in the specified group receive notifications. For more information about how to create a contact group, see Notification contacts.
-
Contact
1. Add one or more contacts to receive notifications. After you click Add contact, click Create contact.
2. On the page that appears, click New contact and enter the contact information.
Contact information includes Name (required), Mobile phone number, Email, Webhook, and Contact group.
-
-
-
Click Complete.
After you create the application, the application list automatically refreshes.
Important-
Application ID: This ID is required for client initialization and serves as the unique identifier for a job group.
-
Application key: This key is required for SDK request authentication. Keep this key confidential.
-
Webhook configuration
To obtain a webhook URL, see the official documentation for DingTalk, WeCom, and Lark.
-
MSE SchedulerX supports WeCom, Lark, and DingTalk webhooks. You can specify multiple webhook URLs by separating them with a comma (,).
-
To receive alerts for DingTalk bots, you must add
SchedulerXas a case-sensitive custom keyword. In the security settings of a DingTalk custom bot, select Custom keywords and add the keywordsMonitoring alertsandSchedulerX. -
If you use a webhook for a contact, ensure that it can correctly parse and process requests in the following format. The example cURL command sends a message to a contact webhook.
curl 'https://oapi.xxxxxxx \ -H 'Content-Type: application/json' \ -d ' { "msgtype": "text", "text": { "content": "SchedulerX alert details" }, "at": { "isAtAll": false } }'
View instances
After connecting your application to MSE SchedulerX, view its running instances to verify the connection:
-
On the Application management page, check the value in the Total instances column.
NoteIf the value in the Total instances column is 0, the application failed to connect.
-
In the Actions column of the application, click View instances.
-
On the View instances page, view the list of workers for the application and the status of each worker.
-
Although a worker typically has only one client, multiple clients are supported.
-
Workers can have a status of Healthy or Busy. By default, jobs are not dispatched to busy workers. In extreme cases, such as when all connected workers are busy, job dispatching may fail. To force jobs to be dispatched to busy workers, in the Edit > Worker busy configuration section, turn the Whether to trigger a busy machine switch to Off.
In the application list, find the application and click Edit in the Actions column.
-
Authorization
You can grant application-level permissions. Both Alibaba Cloud accounts and RAM users can grant access to other RAM users within the same account.
On the Authorization management page, select the target RAM user from the Unauthorized list on the left, and click the >> button to move the user to the Authorized list.
If you see a You do not have the required permissions message when you click Grant, you must use an Alibaba Cloud account to attach the following custom policy to the RAM user:
{
"Version": "1",
"Statement": [
{
"Action": "ram:ListUsers",
"Resource": "*",
"Effect": "Allow"
}
]
}
Delete an application
You can delete an application if you no longer need to create jobs for it. After an application is deleted, all jobs associated with it will no longer be triggered.
-
On the Application management page, find the application that you want to delete and click Delete in the Actions column.
-
In the Delete application dialog box, confirm the details, enter the application name for verification, and then click OK.
-
After you delete an application, its associated jobs are no longer triggered, and it is removed from the application list.
-
If you accidentally delete an application and need to restore it, join the DingTalk group (ID: 23103656) and contact SchedulerX technical support.