Rapid Recovery 6.5 - User Guide

Understanding deduplication cache or dictionary size and storage locations

Global deduplication reduces the amount of disk storage space required for data your Core backs up. Each repository is deduplicated, storing each unique block once physically on disk, and using virtual references or pointers to those blocks in subsequent backups. To identify duplicate blocks, Rapid Recovery includes a deduplication cache for deduplication volume manager (DVM) repositories. The cache holds references to unique blocks.

By default, for DVM repositories, this deduplication cache is 1.5GB. This size is sufficient for many repositories. Until this cache is exceeded, your data is deduplicated across the repository. Once the amount of redundant information is so great that the deduplication cache is full, your repository can no longer take full advantage of further deduplication for newly added data. The amount of data saved in your repository before the deduplication cache fills varies by the type of data being backed up, and is different for every user.

You can increase or decrease the size of the deduplication cache by adjusting the Deduplication cache size value in the DVM Deduplication Cache Core settings. For more information on how to increase the cache size, see the topic Configuring DVM deduplication cache settings.

When you increase the DVM deduplication cache size, there are two factors to consider: disk space and RAM usage.

Disk space. Two copies of the DVM deduplication cache are stored on disk: a primary cache, and a secondary cache which is a parallel copy. Thus, if using the default cache size of 1.5GB for a DVM repository, 3GB of disk storage is used in your system. As you increase the cache size, the amount of disk space used remains proportionally twice the size of the cache. To ensure proper and fault-resistant performance, the Core dynamically changes the priority of these caches. Both are required, the only difference being that the cache designated as primary is saved first.

RAM usage. When the Rapid Recovery Core starts, it loads the deduplication cache to RAM. The size of the cache therefore affects memory usage for your system. The total amount of RAM the Core uses depends on many factors. These factors include which operations are running, the number of users, the number of protected machines, and the size of the deduplication cache. Each operation the Core performs (transfer, replication, rollup, and so on) consumes more RAM. Once an operation is finished, memory consumption decreases accordingly. However, administrators should consider the highest RAM load requirement for efficient operations.

Default settings for the Rapid Recovery Core place the primary cache, secondary cache, and the metadata cache for DVM repositories in the AppRecovery directory. This folder is installed on the Core machine.

Assuming the Rapid Recovery Core is installed on the C drive, these locations are typically as follows:

For conceptual information about deduplication, see Deduplication in Rapid Recovery.

Configuring DVM deduplication cache settings

1. Navigate to the Rapid Recovery Core Console.
2. On the icon bar, click (Settings), and then do one of the following:
   • From the list of Core settings on the left side of the Settings page, click DVM Deduplication Cache.
   • Scroll down on the right side of the Settings page until you can see the DVM Deduplication Cache heading.
3. If you want to restore default DVM deduplication cache settings at any time, do the following:
   1. At the top of the deduplication cache settings area, click Restore Default.

      The Restore Default dialog box appears

   2. Click Yes to confirm the restore.
4. Click the setting you want to change.

   The setting you selected becomes editable.

5. To change individual deduplication cache settings, enter the configuration information as described in the following table.

   Table 21: DVM deduplication cache settings information

   Setting	Description
   Restore Default	This control resets DVM cache locations to system default locations, which are described for each setting.
   Primary cache location	If you want to change the primary cache location for DVM repositories, then in the Primary cache location text box, type the path for a storage location accessible to the Core. The default location is: C:\ProgramData\AppRecovery\RepositoryMetaData\PrimaryCache Since the primary and secondary caches are the same size, collective storage for these two caches requires twice the amount of space as the amount allocated for the deduplication cache size. For example, if you specify the default amount of 1.5GB for the deduplication cache size, ensure that each of the two storage locations have at least 1.5GB. In particular, if both locations belong to the same drive (for example, the C drive), there must be at least 3.0GB of free disk space.
   Secondary cache location	If you want to change the secondary cache location for DVM repositories, then in the Secondary cache location text box, type the path for a storage location accessible to the Core. The default location is: C:\ProgramData\AppRecovery\RepositoryMetaData\SecondaryCache
   Cache metadata location	If you want to change the cache metadata location for DVM repositories, then in the Cache metadata location text box, type the path for a storage location accessible to the Core. The default location is: C:\ProgramData\AppRecovery\RepositoryMetaData\CacheMetadata
   Deduplication cache size (GB)	If you want to change the deduplication cache size for DVM repositories, then in the Deduplication cache size (GB) text box, enter a new amount in gigabytes. NOTE: If you decrease the size of the deduplication cache, the existing contents of the cache are flushed, and the cache is recreated. The default location is: C:\ProgramData\AppRecovery\RepositoryMetaData\SecondaryCache The minimum cache size setting is 1.5GB. Additionally, the cache size cannot exceed 50 percent of the installed RAM.

6. For each setting, when satisfied with your changes, click to save the change and exit edit mode, or click to exit edit mode without saving.

Configuring Replay engine settings

1. Navigate to the Rapid Recovery Core Console.
2. On the icon bar, click (Settings), and then do one of the following:
   • From the list of Core settings on the left side of the Settings page, click Replay Engine.
   • Scroll down on the right side of the Settings page until you can see the Replay Engine heading.
3. Click on the setting you want to change.

   The setting you selected becomes editable.

4. Enter the configuration information as described in the following table.

   Table 22: Replay engine settings information

   Text Box	Description
   IP address	The Core uses this IP address when performing mount and restore for a recovery point, to allow feedback between protected machines and the Сore. The IP address for the Replay engine automatically populates with the IP address of the Core machine. If you manually enter the server IP address, then this value is used in cases where the protected machine cannot resolve the automatically provided IP address. You do not need to set this value manually unless you are having issues with protected machines being able to communicate with the Core.
   Preferred port	Enter a port number or accept the default setting. The default port is 8007. The port is used to specify the communication channel for the Replay engine.
   Port in use	Represents the port that is in use for the Replay engine configuration.
   Allow port auto-assigning	Click for allow for automatic TCP port assignment.
   Admin group	Enter a new name for the administration group. The default name is BUILTIN\Administrators.
   Minimum asynchronous I/O length	Enter a value or choose the default setting. It describes the minimum asynchronous input/output length. The default setting is 65536.
   Read timeout	Enter a read timeout value or choose the default setting. The default setting is 00:05:00.
   Write timeout	Enter a write timeout value or choose the default setting. The default setting is 00:05:00.
   Receive buffer size	Enter an inbound buffer size or accept the default setting. The default setting is 8192.
   Send buffer size	Enter an outbound buffer size or accept the default setting. The default setting is 8192.
   No delay	It is recommended that you leave this check box unchecked as doing otherwise will impact network efficiency. If you determine that you need to modify this setting, contact Quest Data Protection Supportfor guidance in doing so.

5. For each setting, when satisfied with your changes, click to save the change and exit edit mode, or click to exit edit mode without saving.

Configuring deployment settings

1. Navigate to the Rapid Recovery Core Console.
2. On the icon bar, click (Settings), and then do one of the following:
   • From the list of Core settings on the left side of the Settings page, click Deploy.
   • Scroll down on the right side of the Settings page until you can see the Deploy heading.
3. Click on the setting you want to change.

   The setting you selected becomes editable.

4. Enter the configuration information as described in the following table.

   Table 23: Deployment settings information

   Text Box	Description
   Agent installer name	The default filename is Agent-Web.exe. If you wish to change this file name for any reason, you can use this setting to specify a new name of the Core Web Installer executable file. This file streams a download of the latest version of the Rapid Recovery Core installer, which runs directly from the web and lets you pause and resume the process as needed.
   Core address	Enter the address of your Core server. This typically consists of the protocol, the name of your Core server and port, and the directory where the Core files reside. For example, if your server is Sample, this setting is https://sample:8006/apprecovery/admin/Core.
   Failed receive timeout	The amount of time deployment of the Agent software should be attempted before timing out. The default setting is 00:25:00 or twenty-five minutes. If you wish to change this setting, enter the length of time you want the system to attempt to deploy the Agent software before a timeout occurs during read or write events. Uses HH:MM:SS format.
   Maximum parallel installs	This setting controls the maximum number of deployments of the Agent software for the Core to attempt at one time. The default setting is 100.

5. For each setting, when satisfied with your changes, click to save the change and exit edit mode, or click to exit edit mode without saving.