> For the complete documentation index, see [llms.txt](https://asus-isg-aidc.gitbook.io/guide/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://asus-isg-aidc.gitbook.io/guide/latest/portal/overview/faq.md).

# FAQ

| Developer | Last modified |
| --------- | ------------- |
| AIDC Team | 2026/01/23    |

## Q1: How to choose the correct Node/Group?

### Answer

Choosing Node or Group depends on your deployment requirements:

#### When to Choose Group

* Need to install the same service on multiple nodes
* Service requires cluster architecture (e.g., MariaDB Galera, Slurm)
* Want to simplify management by managing a group of nodes uniformly

#### When to Choose Node

* Only need to install on a single node
* Service is Standalone type (e.g., UFM Standalone)
* Need precise control over installation location

### How to Confirm Nodes in a Group

1. Go to **System Overview** or **Deployment Overview**
2. View Group details
3. Confirm the number and status of nodes in the Group

### Common Mistakes

❌ Selecting a Group with only 1 node to deploy cluster services\
❌ Selecting offline or abnormal status nodes\
❌ Selecting the same Group for different roles (e.g., three roles in Slurm)

***

## Q2: How to handle deployment failure?

### Diagnostic Steps

#### 1. Check Error Messages

In Status Monitor:

1. Find the failed task card
2. Expand the card to view details
3. Check error logs or error messages

#### 2. Common Failure Causes and Solutions

| Error Type                     | Possible Cause                       | Solution                        |
| ------------------------------ | ------------------------------------ | ------------------------------- |
| **Network Connection Failure** | Node unreachable                     | Check node network status       |
| **Insufficient Resources**     | Disk space or memory shortage        | Clean up or expand resources    |
| **Permission Issues**          | Insufficient SSH or sudo permissions | Check node permission settings  |
| **Parameter Errors**           | Invalid IP or path                   | Correct parameters and redeploy |
| **Missing Dependencies**       | Required packages missing            | Install necessary dependencies  |
| **Service Conflicts**          | Port already in use                  | Stop conflicting services       |

#### 3. Check Node Status

* Expand the task card
* View execution status of each node
* Identify the failed node

### Retry Deployment

1. Click the **Edit** button on the failed task
2. Check and correct settings
3. Resubmit deployment

***

## Q3: How to redeploy?

### Method 1: Edit Existing Task

Applicable when: Partial settings need modification

1. Find the task in Status Monitor
2. Click the **Edit** button
3. Modify settings
4. Click **Deploy** to redeploy

### Method 2: Create New Deployment

Applicable when: Starting completely fresh

1. Delete the old failed task (optional)
2. Go to Service Catalog
3. Select service and click Deploy
4. Fill in settings again
5. Execute deployment

### Method 3: Modify/ReRun from Failed Task

1. View settings of the failed task
2. Record important parameters
3. Create new deployment
4. Fill in same parameters (correcting the erroneous parts)

***

## Q4: Status update frequency?

### Auto-refresh Mechanism

The system automatically refreshes status information periodically:

| Status         | Refresh Frequency                  |
| -------------- | ---------------------------------- |
| In Progress    | More frequent (about 5-10 seconds) |
| Other statuses | Less frequent (about 30 seconds)   |

### Manual Refresh

To update status immediately:

1. **Refresh page**: F5 or click browser refresh button
2. **Switch tabs**: Switch between Catalog and Status

### Notes

* Frequent refreshing may increase server load
* During deployment, it's recommended to wait for auto-update
* If no update for a long time, it may be a network issue

***

## Q5: Why are some Groups not selectable?

### Possible Causes

1. **Already selected by another role** (e.g., Slurm's mutual exclusion mechanism)
2. **Group status abnormal**
3. **No available nodes in Group**
4. **Permission restrictions**

### Solutions

* Select another available Group
* Confirm Group status is normal
* Check account permissions

***

## Q6: How to view deployment logs?

### In Status

1. Find the corresponding task card
2. Expand the card
3. View execution details and error messages

## Q7: What to do if deployment takes too long?

### Normal Deployment Time Reference

| Service        | Estimated Time |
| -------------- | -------------- |
| Docker Engine  | 5-10 minutes   |
| Podman Engine  | 5-10 minutes   |
| MariaDB Galera | 10-20 minutes  |
| Slurm Cluster  | 15-30 minutes  |
| Native K8s     | 20-45 minutes  |
| NMX-M          | 20-40 minutes  |
| WEKA           | 30-60 minutes  |
| UFM Standalone | 10-15 minutes  |

### Timeout Handling

If significantly exceeding estimated time:

1. Check if task status has changed to Timeout
2. Check node network connection status
3. Check node resource usage
4. Consider canceling and redeploying

***

## Q8: Can multiple services be deployed simultaneously?

### Answer

Yes, the system supports multiple deployment tasks running simultaneously.

### Notes

* Ensure sufficient node resources
* Avoid deploying multiple services on the same node simultaneously
* Monitor the status of each task

***

## Q9: How to confirm service is running successfully?

### Verification Methods

1. **Status shows Completed**
2. **SSH into node to verify**

```bash
# Docker
docker ps

# Kubernetes
kubectl get nodes

# Slurm
sinfo

# MariaDB
mysql -u root -p -e "SHOW STATUS LIKE 'wsrep_%';"
```

3. **Access service Web UI** (if applicable)

***

## Q10: How to cancel an in-progress deployment?

### Current Limitations

Currently, the system does not support directly canceling in-progress deployment tasks.

### Alternatives

1. Wait for task to complete or timeout
2. Delete after task completion

***

## Need More Help?

If the above questions don't solve your problem, please:

1. Check [Technical Documentation](https://gitlab.com/AIDCTeam/gitbook-docs-portal/-/tree/main/projects/AIDC/guide/Portal/Marketplace/technical-docs.md)
2. Contact system administrator
3. Submit a technical support ticket

***

## Next Steps

👉 [Technical Documentation](https://gitlab.com/AIDCTeam/gitbook-docs-portal/-/tree/main/projects/AIDC/guide/Portal/Marketplace/technical-docs.md) - Developer technical documentation