InfoScale™ 9.0 Storage Foundation Administrator's Guide - Linux

Ease of use:

A faster recovery process involving a few clicks ensures a better Recovery Time Objective (RTO). Single Click Recovery supports data recovery of at least one hour thereby supporting an improved Recovery Point Objective (RPO).

Checkpoints management and validation:

The vxschadm command has new options for listing and validating immutable SecureFS checkpoints, their status, integrity, and required space.

To display a list of checkpoints, run the following command:

#/opt/VRTS/bin/vxschadm show [checkpoint] <mntpt>

This command displays all the SecureFS checkpoints that have names starting with "secfs_" prefix, including their retention periods and flags. You can use the information to verify which checkpoints are available for recovery.

To validate a checkpoint, run the following command:

#/opt/VRTS/bin/vxschadm validate [checkpoint] [-a application_name] <mntpt>

This command verifies if the checkpoint is present on all the application-related mount points. If these details are not available in the SecureFS config file, it verifies this information from the main database mount point. The command also performs a quick mount and unmount operation to ensure that the checkpoint can be accessed and is intact. Finally, the command verifies that there is enough free disk space available to perform the recovery.

Ensure that you have enough free disk space available or provisioned before you initiate a recovery. The general guideline is to have roughly twice the current used space. For example, a 500 GB checkpoint requires a minimum of an additional 500 GB of free space on the main file system.

Automated and safe recovery:

The recover command not only initiates the restore process, but also creates a temporary checkpoint to revert to in case the restore operation fails.

To initiate a restore, run the following command:

#/opt/VRTS/bin/vxschadm recover [checkpoint] <ckptname> [-a application_name] <mntpt>

This command repeats the checkpoint validation steps that are mentioned earlier (ensures that the checkpoint exists on every required mount point, is accessible, and there is enough storage disk space available). The source checkpoint that is used for recovery is marked as non-removable during the recovery to prevent any accidental deletions, for example in case of ENOSPC where the disk space is low.

Additionally, this command creates a temporary checkpoint with a prefix "secfs_TIMESTAMP_temp" for each mount point. This temporary checkpoint acts as a fallback in case the recovery fails, the system can revert to the original state. The temporary checkpoint is deleted after a successful recovery.

The command also verifies that SecureFS is disabled so that no new changes are made and also checks that the Cluster Sever (VCS) resources are frozen. Note that disabling SecureFS and freezing VCS resources is done from VIOM.

Parallel processing:

The recovery process runs in parallel if there are multiple mount points to improve efficiency. The vxschadm utility determines the optimal number of parallel threads depending on the CPU count or the file systems involved. Any additional tasks are automatically queued and their status shows as "About to start..".

Application-level integration:

The recovery process enables interaction with application-level functions such as database shutdown and startup, thereby ensuring that the application state remains consistent throughout the restore operation. Application-aware recovery is supported for Oracle and PostgreSQL applications.

Interactive monitoring:

The VIOM GUI provides continuous updates on the restore process ensuring that you are aware of the status of the recovery.