Overview
GooseFS organizes and manages cached data through namespaces. A GooseFS namespace is a unified access interface for the GooseFS cluster to access data, and it can mount various underlying storage systems, including COS and CHDFS. This page is used to configure the basic properties of a namespace and its metadata discovery capability.
Namespace Basic Configuration
Namespace management supports creating, updating, and deleting namespaces, associating a COS storage bucket with a GooseFS cluster, and setting a cache read/write policy.
2. In the left navigation, select GooseFS > Instance List to enter the GooseFS cluster list page.
3. Select the GooseFS cluster where you need to create a namespace, enter the cluster details page, click Namespace in the sidebar, and enter the namespace subpage.
4. On the namespace page, click Add Namespace and fill in the following fields.
Bucket Source: Selectable from your account or other accounts' buckets.
Notes:
When selecting buckets from other accounts, ensure that the corresponding account has read/write permissions for the corresponding bucket.
COS Bucket: Select the bucket you need to mount.
Project Name: Namespace name.
Notes:
A namespace name cannot be modified after creation.
Mounting Range: Selectable for the entire bucket or a specified directory prefix.
UFS URI: An address used to access and operate files in the UFS file system.
Read Policy: Used to manage the cache policy of data in the GooseFS cluster when reading data from GooseFS. Currently only supports CACHE.
Write Policy: Used to manage the cache policy of data in the GooseFS cluster when data is written through GooseFS. The available configuration options are described as follows:
Dual-write to GooseFS and COS: Write requests simultaneously write data to both GooseFS and COS.
Direct Write to COS: Write requests directly write data to COS without going through GooseFS. This configuration may cause inconsistency between GooseFS metadata and COS.
UFS Attributes: Used to set configuration information for accessing underlying storage, such as access keys.
fs.cosn.bucket.region: Selected bucket region, such as ap-beijing.
fs.cosn.flush.enabled: Whether to enable the data refresh mechanism, the default is true.
fs.cosn.upload.part.size: The shard size for multi-part upload, default is 8MB.
fs.cosn.userinfo.secretId: used to mark the identity of the API caller. One APPID can create multiple cloud API keys.
fs.cosn.userinfo.secretKey: Encryption signature key for your Tencent Cloud account, used to verify request legitimacy.
Notes:
SecretId and SecretKey are collectively referred to as TencentCloud API keys, which are security credentials required for identity verification when users access Tencent Cloud APIs. Create them on the CAM page. In the left sidebar, choose Access Key > API Key Management, and click Create Key to obtain them. 5. After confirming that the above information is correct, click OK to create the namespace. After the namespace is created, you can view it in the namespace list and manage it in the operation column on the right, such as viewing details, updating, or deleting the namespace.
View Details: View namespace details.
Update operation: Supports modifying Write Policy and UFS Attributes.
Delete operation: Supports deleting the current namespace. Note that after the namespace is deleted, your cached data and services will be interrupted. Confirm that you no longer use this namespace before deleting it.
Metadata Discovery
Metadata Discovery is an active metadata synchronization service launched by GooseFS. It automatically resolves consistency issues between the cache layer and the underlying storage, relieving the pressure of manual metadata preheating on the client side. You can specify specific COS events in the Metadata Discovery configuration, and GooseFS actively syncs the events to the cache layer. Metadata Discovery enables near-real-time metadata synchronization. Under high QPS COS request loads, metadata changes can be synced to GooseFS within minutes.
Notes:
Metadata Discovery is currently a beta feature and is free of charge during the beta period. To apply for beta access, please contact us. Metadata Discovery requires GooseFS cluster version 1.5.1 or later. If you need to upgrade your cluster version, please contact us. Only GooseFS clusters in Fully Managed Hosting Mode and Master-Only Hosting Mode support enabling Metadata Discovery.
If versioning has been enabled for your COS bucket, the behavior in overwrite and delete scenarios is as follows:
Overwriting uploaded files: Metadata Discovery can synchronize normally, syncing the metadata of each file write to GooseFS.
Delete: The current version of GooseFS does not support syncing delete events to GooseFS.
The timeliness of Metadata Discovery depends on the specific business model. For the actual sync time, refer to the online performance.
You can enable Metadata Discovery by following these steps:
2. In the left navigation, select GooseFS > Instance List to enter the GooseFS cluster list page.
3. Select the GooseFS cluster where you need to create a namespace, enter the cluster details page, click Namespace in the sidebar, and enter the namespace subpage.
4. On the namespace page, click Add Namespace, and fill in the following fields in the pop-up window.
|
Metadata discovery capability switch | Open |
Event types to sync to GooseFS. Supported options: Select all. Select by event type. Triggered after an object is uploaded by any method. Triggered only after the object content is deleted. Select by specific event. Triggered only after an object is uploaded through the PUT Object API. Triggered only after an object is uploaded through the POST Object API. Triggered only after an object is uploaded through the PUT Object - Copy API. Triggered only after an object is uploaded through the Complete Multipart Upload API. | Select as needed, for example, select all. |
Namespace range that triggers metadata discovery. Options include the mount range of the entire namespace or a subdirectory under the mount range of the namespace. | Select as needed, for example, the mount range of the namespace. |
Performs upload event reconciliation on a daily basis. After this feature is enabled, a temporary inventory is delivered to the COS bucket every day to verify the metadata discovery records from the previous day. If any upload events are missed, they are supplemented. The temporary inventory is deleted within 48 hours after reconciliation is completed. Notes: This feature requires GooseFS cluster version 1.5.2 or later. If you need to upgrade your cluster version, please contact us. | Disabled. |
This path is used to deliver temporary COS inventories. | Configure buckets as needed. To avoid circular triggering, do not use buckets already mounted in the namespace. |
Delivery path of reconciliation inventories in the COS bucket. Notes: To ensure normal reconciliation, we do not recommend that you perform operations on files in this directory after the configuration is complete. Temporary lists will be deleted within 48 hours after reconciliation is complete. | Fill in the bucket path as needed. To ensure reconciliation runs properly, create a separate path for the reconciliation inventory. |
If you need to modify the Metadata Discovery configuration, you can click Update for the configured namespace on the namespace list page to re-edit the configuration or disable the namespace. Note that when you disable Metadata Discovery or delete a namespace, GooseFS immediately stops metadata synchronization. If there are still COS requests that have not been synced to GooseFS, these request events may be lost. Therefore, before disabling Metadata Discovery, make sure that the required metadata has been synced to GooseFS.
Following Steps
After creating a namespace, you can manage it on your nodes by adding new clients. For details, see client management.