The cache feature is enabled by default on a Hybrid Backup Recovery (HBR) client. You can use a cache to accelerate the data backup process. However, you cannot use the cache to help restore data on a local or remote host. This topic describes how to optimize the settings of a cache.

Prerequisites

An HBR client is installed for local files. The version of the client is 1.5.0 or later. For more information, see Prepare for a data backup.

Background information

The cache feature allows you to accelerate the data backup process by caching data entry IDs and metadata. The feature reduces network requests during data backup. You can disable this feature or optimize its settings on the backup source.

Procedure

You can perform the following steps to create a cache configuration file. You can also disable the cache feature or optimize its settings by using the configuration file. This file is optional. If you do not create a file, the default settings of the feature are used. The backup process is not affected.

  1. Log on to the on-premises server or virtual machine (VM) whose files you want to back up.
  2. Open the installation folder of the HBR client.
    You can find the installation path based on the following default installation paths.
    • For backup clients of an earlier version

      Linux: /opt/alibabacloud/hbr

      Windows: C:\Program Files\Aliyun Hybrid Backup Service

    • Backup clients of the latest version

      Linux: /opt/alibabacloud/hbrclient

      Windows: C:\Program Files\Aliyun Hybrid Backup Service Client

  3. In a subdirectory of the client folder, create a file named hbr.config, and add the information of the data entry ID and metadata cache to the file.
    Note The hbr.config file must be at the same directory level as the ids file.
    The following example shows the configurations of the hbr.config file:
    disable_blob_cache = false
    max_blob_cache_weight = 0.15
    cache_prefix = D:\CacheFolder
    max_retain_count = 16 
    
    disable_file_cache = false
    file_cache_max_size_hint = 2GB           
    Parameter Description
    disable_blob_cache Specifies whether to cache data entry IDs. Valid values:
    • true: Data entry IDs are cached.
    • false: Data entry IDs are not cached.
    max_blob_cache_weight The maximum system memory that can be used by the cache for data entry IDs. Default value: 0.15. The value indicates 15% of the total system memory. The value must be between 0 and 1.
    cache_prefix The path in which the cached data entry IDs are stored. The value must be an absolute path.
    max_retain_count The maximum number of cached data entry IDs.
    disable_file_cache Specifies whether to cache metadata. Valid values:
    • true: Metadata is cached.
    • false: Metadata is not cached.
    file_cache_max_size_hint The maximum size of a metadata cache file. The actual size may exceed the specified value. Default value: 2 GB.
    Note
    • If you set the parameter to 2 GB, you can back up a minimum of 4 TB of metadata.
    • If you set the parameter to a small value, the backup does not fail but the cache performance is degraded.
    • The value of the parameter cannot exceed the available disk space.
    Notice
    • The parameter settings take effect immediately after you save the hbr.config file. You do not need to restart the server or VM.
    • The parameter settings are applied only to subsequent backup jobs. You cannot accelerate existing backup jobs by caching data entry IDs and metadata.