NetApp SANtricity Cloud Connector 3.1

Transcription

1 Technical Report NetApp SANtricity Cloud Connector 3.1 Robert Stankey, Janelle Sperry, NetApp June 2018 TR-4658 In partnership with Abstract NetApp SANtricity Cloud Connector 3.1 is a host-based Linux application built specifically for engagement with the Data Fabric through NetApp E-Series storage systems and arrays. NetApp has partnered with Amazon Web Services to allow full-block and file-based backup and recovery of E-Series volumes to Amazon S3, NetApp StorageGRID storage, and NetApp Cloud Backup storage. Available for installation on Red Hat and SUSE Linux platforms, SANtricity Cloud Connector is a packaged solution (.bin file). After you install SANtricity Cloud Connector, you can configure it to perform backup and restore jobs for E- Series volumes to your existing Amazon S3, StorageGRID, or NetApp Cloud Backup account. All jobs performed through SANtricity Cloud Connector use RESTful APIs. This report provides the technical details for each feature, general limitations, and specific restrictions that you must consider to take advantage of its features.

2 Table of Contents 1 Introduction Overview Terminology Workflow Diagram Types of Backup...7 Image-Based Backup... 7 File-Based Backup System Requirements Host Hardware Requirements Compatible Storage Arrays Compatible Operating Systems Using SANtricity Cloud Connector Installation of SANtricity Cloud Connector Installation of Device Mapper Multipathing Prerequisites for Using SANtricity Cloud Connector Initial Configuration of SANtricity Cloud Connector Performing an Image-Based Backup Performing an Image-Based Restore Performing File-Based Backup Performing File-Based Restore Benchmark Results Overview Image-Based Backup and Restore Results File-Based Backup and Restore Results Conclusions from the Benchmark Test Summary Appendix A: E-Series Snapshot Images NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

3 Snapshot Image Snapshot Volume Consistency Group Rollback I/O Handling to Snapshot Volumes Snapshot COW Technology Write Requests Read Requests Appendix B: Manual Procedures for Image-Based Backup/Restore B.1 Performing an Image-Based Backup Verify Backup Snapshot Volume Is Visible to SANtricity Cloud Connector Host Create a Backup Job Remove Backup Snapshot Volume from SANtricity Cloud Connector Host B.2 Performing an Image-Based Restore Create a Restore Job Remove the Restore Volume from the SANtricity Cloud Connector Host Appendix C: Manual Procedures for File-Based Backup and Restore C.1 Performing File-Based Restore Verify SANtricity Cloud Connector Host Has Visibility to the File System Create a Backup Job Remove Backup Snapshot Volume from SANtricity Cloud Connector Host C.2 Performing File-Based Restore Verify SANtricity Cloud Connector Host Has Visibility to the Restore Volume Create a Restore Job Remove the Restore Volume from the SCC Host Where to Find Additional Information Version History LIST OF TABLES Table 1) Operating system requirements NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

4 Table 2) Image-based backup and restore results Table 3) File-based backup and restore results Table 4) I/O handling to a Snapshot base volume LIST OF FIGURES Figure 1) Workflow diagram....6 Figure 2) Cloud Connector performance demonstration Figure 3) Snapshot image and Snapshot group Figure 4) Snapshot volume Figure 5) Consistency group Figure 6) Rollback from Snapshot image to base volume Figure 7) Snapshot COW technology Figure 8) COW example Figure 9) WWN of restore volume Figure 10) WWN of Snapshot volume NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

5 1 Introduction NetApp SANtricity Cloud Connector is a host-based Linux application that allows you to perform full-block and file-based backup and recovery of NetApp E-Series volumes to the Amazon Simple Storage Service (S3), NetApp StorageGRID storage, and NetApp Cloud Backup storage. Available for installation on Red Hat and SUSE Linux platforms, SANtricity Cloud Connector is a packaged solution (.bin file). After it is installed, you can configure SANtricity Cloud Connector to perform backup and restore jobs for E-Series volumes to your existing Amazon S3, StorageGRID, or Cloud Backup account. All jobs performed through SANtricity Cloud Connector use RESTful APIs. 2 Overview The following sections introduce the terminology used in SANtricity Cloud Connector discussions and operations. They also provide a detailed workflow diagram for running applications to various cloud targets, and they define the types of backups supported by the SANtricity Cloud Connector application. 2.1 Terminology The following terms are used throughout this document, in the SANtricity GUIs, and in the storage and data management industry. SANtricity Cloud Connector host. The Linux host on which SANtricity Cloud Connector is installed. Volume or base volume. An E-Series volume that is typically mapped to a host system through LUN mapping. Partition. A host term that defines a way to divide a volume into multiple sections and then treats each section independently from the others. File system. A host term that defines and controls how data is stored and retrieved from a partition. For example, ext4 is a type of file system common with Linux; it lets you organize the data in a hierarchy of directories and files. Mount point. A host term that defines a directory in a currently accessible file system on which an additional file system can be logically attached (mounted). NetApp Snapshot image. A logical point-in-time (created at a specific moment) image of the content of a base volume. A NetApp SANtricity Snapshot image is not directly read/write accessible to hosts. Snapshot group. A collection of Snapshot images of a single base volume. Snapshot volume. A standard volume that allows the host to access the Snapshot image in a Snapshot group. Backup volume or backup base volume. The volume or base volume selected for a backup. Backup Snapshot image. A Snapshot image that has been created for use in a SANtricity Cloud Connector based backup. Backup Snapshot volume. A standard volume that allows the SANtricity Cloud Connector host to access the backup Snapshot image. Backup target. The storage destination for the backed-up data. This target is either an S3 bucket or an NFS mount point based on the intended destination. Restore source. The S3 bucket or NFS mount point from which data is restored. Restore volume or restore base volume. The volume or base volume selected to receive the restored data. 5 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

6 2.2 Workflow Diagram Figure 1 depicts a high-level Cloud Connector workflow diagram in a data center environment that uses E-Series storage for three common use cases: , databases, and a virtualization environment. The diagram shows activities and interfaces used to push data to and from multiple public and private cloud targets. Figure 1) Workflow diagram. The following list describes each numbered item in Figure 1. Customer production environment. A DAS or SAN environment with E-Series storage arrays. Cloud Connector uses Snapshot volumes for the backup operation. Before you request a backup, you must manually quiesce any application that uses the base volumes to ensure data consistency. The backup process creates the Snapshot volume and maps it to the Cloud Connector server through the host interconnect. Base volumes for the restore operation. During a restore operation, base volumes are mapped to the Cloud Connector server through the host interconnect. The restore data files are read from the restore source to reconstruct the restore volume. SANtricity Cloud Connector. A host-based Linux application that allows you to back up and restore E-Series volumes. All backup and restore jobs use RESTful API calls to the application. Backup 6 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

7 operations consist of reading data from the Snapshot volumes and copying the data to one of several destinations: Note: NetApp Cloud Backup, through an NFS mount, which can then be transferred to Amazon Web Services (AWS) with the HTTPS and S3 protocols. AWS, with the HTTPS and S3 protocols. StorageGRID, with the HTTPS and S3 protocols. The restore operation consists of reading the data from the restore source and reconstructing the data to the restore volumes. NetApp SANtricity Web Services Proxy. Used by SANtricity Cloud Connector to communicate with the E-Series array. A Cloud Backup virtual or physical appliance. The StorageGRID solution. A bucket for an AWS cloud account. 2.3 Types of Backup SANtricity Cloud Connector supports two types of backup: image-based backups and file-based backups. Both types are described in this section. Image-Based Backup An image-based backup reads the raw data blocks from a Snapshot volume and backs them up to a file known as an image. All data blocks on the Snapshot volume are backed up, including empty blocks, blocks occupied by deleted files, blocks associated with partitioning, and file system metadata. Image backups have the advantage of storing all information from the Snapshot volume regardless of the partitioning scheme or file systems on it. The image is not stored on the backup target as a single file. Rather it is broken up into a series of 64MB data chunks. The data chunks enable SANtricity Cloud Connector to use multiple connections to the backup target, improving performance. For backups to StorageGRID and AWS, each data chunk uses a separate encryption key for encryption. The key is a SHA256 hash consisting of both a user-supplied passphrase and the SHA256 hash of the user data. For backups to NetApp Cloud Backup, SANtricity Cloud Connector does not encrypt the data chunks because NetApp Cloud Backup performs this operation. File-Based Backup A file-based backup reads the files in a file system partition and backs them up into a series of 64MB data chunks. A file-based backup does not back up delete files or partitioning and file system metadata. As with image-based backups, the data chunks allow SANtricity Cloud Connector to use multiple connections to the backup target, improving performance. For backups to StorageGRID and AWS, each data chunk uses a separate encryption key to encrypt the chunk. The key is a SHA256 hash consisting of both a user-supplied passphrase and the SHA256 hash of the user data. For backups to NetApp Cloud Backup, SANtricity Cloud Connector does not encrypt the data chunks, because NetApp Cloud Backup performs this operation. 7 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

8 3 System Requirements 3.1 Host Hardware Requirements Before installing NetApp SANtricity Cloud Connector software, verify that your system meets the following host hardware requirements: At least 5GB of memory; 4GB for the maximum configured heap size At least 250MB of free disk space for the software installation 3.2 Compatible Storage Arrays For a complete and up-to-date listing of all compatible storage arrays and firmware for SANtricity Cloud Connector, refer to the NetApp Interoperability Matrix Tool. 3.3 Compatible Operating Systems The SANtricity Cloud Connector 3.1 application is compatible with and supported on the operating systems shown in Table 1. Table 1) Operating system requirements. Operating System Version Architecture Red Hat Enterprise Linux (RHEL) 7.x 64 bit SUSE Linux Enterprise Server (SLES) 12.x 64 bit 4 Using SANtricity Cloud Connector This technical report provides SANtricity Cloud Connector setup and configuration steps for backing up and restoring E-Series volumes to an S3-compatible target such as Amazon S Installation of SANtricity Cloud Connector The Cloud Connector packaged solution (.bin file) is available for Red Hat and SUSE Linux platforms only. During the installation process, you must specify the listening port number for SANtricity Cloud Connector. After it is installed, SANtricity Cloud Connector runs as a daemon process. Note: SANtricity Cloud Connector and your current web services proxy must be installed at different port locations. To download SANtricity Cloud Connector, go to the NetApp Support Site Software Download page, reference E-Series SANtricity Cloud Connector, and follow the installation instructions in the SANtricity Cloud Connector User Guide. 4.2 Installation of Device Mapper Multipathing Installation of Device Mapper Multipathing (DM-Multipath) allows backup and restore operations to continue if there is a loss of access to a controller. For information about how to set up and configure Device Mapper, refer to the SANtricity Storage Manager Multipath Drivers Guide for the SANtricity release you are using. 4.3 Prerequisites for Using SANtricity Cloud Connector Verify that your system meets these basic prerequisites before you use Cloud Connector software: 8 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

9 SANtricity Web Services Proxy has been installed on the SANtricity Cloud Connector host. For more information about installing SANtricity Web Services Proxy, see the NetApp SANtricity Web Services Proxy 2.11 Installation Guide. SANtricity Web Services Proxy has been configured to use port SANtricity Cloud Connector has been configured to use port There is at least one host connection to each controller from the SANtricity Cloud Connector host. Linux Device Mapper Multipathing (DM-Multipath) has been installed and configured on the Linux server. The E-Series array has a disk pool (for example, SccWalkthru_Disk_Pool_1-2 TB) with several volumes: Volume name: UserData_Vol_1 (500GB) Volume name: RestoreData_Vol_1 (500GB) 5 Initial Configuration of SANtricity Cloud Connector The following procedure describes the initial configuration of Cloud Connector. To perform initial configuration of SANtricity Cloud Connector, use a supported browser to connect to the server (for example, Use the default password of password at the login screen. After you log in, a configuration wizard guides you through the process of providing settings required for SANtricity Cloud Connector. 9 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

10 Provide a new administrator password to use for subsequent logins and click Next. 10 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

11 Specify connection and login information for the web services proxy used to connect with E-Series storage. Click Next. Storage arrays must first be added to the web services proxy to use them with Cloud Connector. This screen shows all arrays currently known to the web services proxy to which you have connected. Select the arrays with the volumes that you want to back up. Click Next. 11 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

12 For each storage array selected in the previous step, select the host mapping to the Cloud Connector host. After you are finished, click Next. Specify connection information for the backup target. Click Next. Specify the pass phrase to be used as part of the encryption key for the data chunks. Click Next. 12 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

13 SANtricity Cloud Connector validates the required settings and provides a summary review. Click Finish to exit the initial configuration wizard. You then see the backup summary screen. 13 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

14 6 Performing an Image-Based Backup The following steps are required for an image-based backup. In this example, you back up all the data from UserData_Vol_1. Use these steps whether the backup target is AWS, NetApp StorageGRID, or NetApp Cloud Backup. From the Backup tab, click Create to create a new backup job. Select the image-based backup option, enter a name for the job, and click Next. Cloud Connector automatically determines which volumes are eligible for backup and presents them in the next step. 14 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

15 Select the Volume you want to back up from the Create Backup Job dialog box and click Finish. Cloud Connector automatically creates a Snapshot copy of the volume and maps it to the Cloud Connector host. Note: You are asked whether you want to create another backup. Skip the creation of another backup. 15 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

16 The backup job summary screen shows you the status of all jobs. 16 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

17 7 Performing an Image-Based Restore The following steps are required for an image-based restore. In this example, you restore all the data to RestoreData_Vol_1. Use these steps whether the restore source is AWS, NetApp StorageGRID, or NetApp Cloud Backup. From the Restore tab, click Create to create a new restore job. Use the backup job created in the section Performing an Image-Based Backup and click Next. Cloud Connector automatically determines which volumes are eligible for restore and presents them in the next step. 17 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

18 Select the restore volume and then click Finish. Cloud Connector automatically maps the selected volume to the Cloud Connector host. 18 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

19 The restore job summary screen now shows the status of the restore job you just scheduled. 19 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

20 8 Performing File-Based Backup The following steps are required for a file-based backup. In this example, UserData_Vol_1 contains a single Linux ext4 file system on it. Use these steps whether the backup target is AWS, NetApp StorageGRID, or NetApp Cloud Backup. Note: A file-based backup unconditionally backs up all files on the file system you specify. However, you can perform a selective restore of files and folders. From the Backup tab, click Create to create a new backup job. Select the file-based backup option, enter a name for the job, and click Next. 20 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

21 From the storage array, select the volume containing the file system you want to back up. Then click Next. 21 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

22 SANtricity Cloud Connector automatically maps the selected volume to the Cloud Connector host and mounts any file systems that are on it. In the Create Backup Job dialog box, select the file system from which you want to back up. Then click Finish. 22 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

23 After you select the file system for backup and click Finish, you are asked whether you want to create another backup. Skip creation of another backup. The backup job summary in the following image shows you the status of all jobs. 9 Performing File-Based Restore The following steps are required for a file-based restore. In this example, you restore the contents of several folders to a partition that has been created on a restore volume (RestoreData_Vol_1). Use these steps whether the restore source is AWS, NetApp StorageGRID, or NetApp Cloud Backup. From the Restore tab, click Create to create a new restore job. From the Create Restore Job dialog box, select the previously created file-based backup. Click Next. 23 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

24 Next, you are asked to specify the files you want restored. In this example, the /bin folder has been selected, causing all folders and files therein to be checked. Click Next. 24 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

25 Next, you are asked to select the restore volume to use for the restore. After you have chosen the volume, click Next. Cloud Connector automatically maps the selected volume to the Cloud Connector host and mounts any file systems on it. 25 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

26 Next, you are asked to select the restore volume to use for the restore. After you have chosen the volume, click Next. 26 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

27 The restore job summary screen now shows the status of the restore job you just scheduled. After the restore has completed, Cloud Connector automatically unmounts the file systems and unmaps the volume from the Cloud Connector host. 27 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

28 10 Benchmark Results 10.1 Overview Figure 2 illustrates a configuration used to demonstrate Cloud Connector performance between an E- Series array and AWS. Figure 2) Cloud Connector performance demonstration. The following components were included in this test system: An E-Series E2800 using 8GB of memory per controller was chosen as the storage array. The drives used in the storage system were SAS HDD 10K RPM 12GBps, and the firmware was Three volumes were created from a five-drive RAID 6 configuration with DA enabled and no encryption. The sizes chosen for the three volumes were 5GB, 50GB, and 500GB. The interconnect between the E-Series E2800 and the SANtricity Cloud Connector host was 10Gbps iscsi. The internal network infrastructure between the SANtricity Cloud Connector host and firewall to the external internet was 1GB. We used Cloud Harmony s speed test for AWS to help gauge the latency of the network between the SANtricity Cloud Connector host and AWS. 28 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

29 10.2 Image-Based Backup and Restore Results The benchmarks presented in Table 2 represent an image-based backup and restore scenario. In these tests, the whole volume s capacity is written with random data. This situation represents a worst-case backup scenario because it restricts Cloud Connector s ability to deduplicate data prior to backup. Table 2) Image-based backup and restore results. Host to Target Description Backup Restore Cloud Connector host to and from storage (10Gb iscsi) 5GB 50GB 500GB 5GB 50GB 500GB Rx Mibps Tx Mibps Total throughput Cloud Connector host to and from AWS (average latency 73.5ms) Rx Mibps Tx Mibps Total throughput Total transfer time (seconds) 116 1,029 11, ,947 52,440 Cloud Connector host 1 utilization CPU % average Memory % average E-Series E utilization IOPS average MiBps average CPU % average Cloud Connector host: A Dell 720 server with two 2.0GHz six-core Xeon processors (E5-2620) and 80GB of DDR3 single-bit ECC 1600MHz memory. The network interface consists of one 10G Ethernet connection to the array and a 1Gb connection to the infrastructure. 2 E-Series E2800 with 8GB memory per controller. RAID 6 (5 drives), DA enabled, and no encryption. SAS HHD 10K rpm 12Gbps. Firmware level File-Based Backup and Restore Results These benchmarks represent a file-based backup and restore scenario. In these tests, we formatted the volumes with the ext4 file system, and we wrote directories and files with random data. Table 3) File-based backup and restore results. Host to Target Description Backup Restore Cloud Connector host to and from storage (10Gb iscsi) 5GB 50GB 500GB 5GB 50GB 500GB 29 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

30 Host to Target Description Backup Restore Rx Mibps Tx Mibps Total throughput Cloud Connector host to and from AWS (average latency 73.5ms) Rx Mibps Tx Mibps Total throughput Total transfer time (seconds) 156 1,395 15, ,987 53,536 Cloud Connector host 1 utilization CPU % average Memory % average E-Series E utilization IOPS average MiBps average CPU % average Cloud Connector host: A Dell 720 server with two 2.0GHz six-core Xeon processors (E5-2620) and 80GB of DDR3 single-bit ECC 1600MHz memory. The network interface consists of one 10G Ethernet connection to the array and a 1Gb connection to the infrastructure. 2 E-series E2800 with 8GB memory per controller. RAID 6 (5 drives), DA enabled, and no encryption. SAS HHD 10K rpm 12Gbps. Firmware level Conclusions from the Benchmark Test Backup and restore operations are very consistent between the volume sizes. The effect on the Cloud Connector host and the E-Series storage array is minimal. The backup operation uses multithreaded processes to help with performance. Unfortunately, the restore operation is currently restricted to a single thread. These results show that the image-based backup and restore performs better than the file-based backup and restore. However, keep in mind that the table is showing worst-case scenarios. Also keep in mind that an image-based restore must restore the full contents of the volume, whereas a filebased restore provides the option to only restore a subset of the data. Connection bandwidth can be thought of in two ways: The bandwidth of the network connection that you have to the backup target itself. A typical business might have 100Mb or 1GbE connections for its internal network. However, connecting to an external backup target (such as Amazon S3) over consumer-grade internet service provides only a fraction of the internal network speed. Faster internet speeds might not be possible, depending on cost and customer location. This situation affects both backup and restore operations. 30 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

31 Bandwidth limits imposed by the backup target service. Services such as Amazon S3 can offer different levels of service based on connection speeds. This situation affects both backup and restore operations. Array-level performance. The SANtricity Cloud Connector application relies on the underlying E- Series and EF-Series array firmware to perform the following tasks: The initial Snapshot copy of the base volume The discovery of the SANtricity Cloud Connector host interface and multipath driver The designation of base Snapshot copy volumes (LUNs) to the SANtricity Cloud Connector host These operations might take place while the storage array is serving other workloads, so the storage array Snapshot feature performance overhead and time to complete a Snapshot copy can vary. The SANtricity Cloud Connector application does not change the performance profile associated with normal SANtricity Snapshot feature functionality. As a result, deciding when to initiate Snapshot copies includes pausing all write I/O to the target volume and some consideration for other workloads the array is simultaneously serving. For example, to minimize the performance overhead associated with initial Snapshot copies, consider taking Snapshot copies during the storage array s lowest overall I/O processing window. Summary SANtricity Cloud Connector 3.1 provides an easy-to-use interface that integrates the SANtricity OS Snapshot feature with the ability to back up and restore data to and from StorageGRID and AWS cloud storage targets. The application runs on a standard Linux host with FC access to the E-Series or EF- Series array. It does not require special skills or training beyond normal array management skills to set up and manage backups to the cloud. This application provides customers who have not invested in cloud integration software with an integrated software alternative to begin their cloud integration journey. The SANtricity Cloud Connector 3.1 application is available for download by users with an active NetApp Customer Success Services site account that includes access to the software download page. Select E- Series SANtricity Cloud Connector from the downloads list. Appendix A: E-Series Snapshot Images NetApp SANtricity Cloud Connector 3.1 creates NetApp SANtricity Snapshot images for backup purposes. After the backup has completed, the Snapshot image is deleted. This section describes what Snapshot images do. For more information, see the E-Series Snapshot Quick Start Guide or Deploying NetApp E-Series and EF-Series Copy Services with Oracle and SQL Server Databases. A Snapshot image is a point-in-time logical copy of a user volume or a set of volumes. You can use a Snapshot image to create a read/write volume that: Is accessible from the host Requires only a small percentage of the base volume capacity Allows rollback of the base volume to restore its data to a previous point in time Snapshot images can also be used for testing and training and for protection against user error and data corruption. Snapshot technology is widely used to test applications and to test upgrades, application patches, OS upgrades, and OS patches before corporate rollout. It is possible to create an identical Snapshot volume to support an in-depth test environment. 31 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

32 Also, during code and application development, you can create parallel environments to speed development of multiple solutions or competing solutions. By using Snapshot volumes for testing, you can create a virtual clone of a volume without having to touch the actual volume itself. By using consistency groups, you can consistently back up applications that are spread across multiple volumes, such as a database, for use in a secondary environment. To maintain the point-in-time image of the base volume, Snapshot images use a repository to keep the original data blocks for blocks that change in the base volume after the image is created. The repository is automatically created when the Snapshot group is created and is, by default, 40% of the size of the base volume. Snapshot images are logical copies of the base volume, so they do not take any capacity from the storage system other than the capacity that the repository requires. Snapshot Image Figure 3 represents how Snapshot images are created from a base volume at specific points in time. Figure 3) Snapshot image and Snapshot group. Snapshot images are collected in a Snapshot group that is associated only with a unique base volume. If there is no Snapshot group when the first Snapshot image of the base volume is created, the storage management software automatically also creates a Snapshot group for the image. The first Snapshot image is an exact logical copy of the base volume at the point in time when the image was created. Thereafter, when the system generates another Snapshot image, it captures only the data changes that occurred after the previous Snapshot image was made. Each Snapshot group includes a Snapshot group repository in which to save Snapshot images. Every Snapshot image is created in the context of exactly one Snapshot group. Up to four Snapshot groups can exist per base volume. When a base volume is in a consistency group, the consistency group is counted toward the limit of four Snapshot groups. Each Snapshot group can in turn contain 32 Snapshot images. As a storage administrator, you can create multiple Snapshot groups to support different schedules for creating Snapshot images or to support more than 32 Snapshot images for the base volume. 32 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

33 The base volume that is associated with a Snapshot group can reside in a disk pool or in a volume group. If the base volume resides in a volume group, the repository members for any associated Snapshot group can reside in a disk pool or in a volume group. If, however, the base volume resides in a disk pool, all repository members for any associated Snapshot group must reside in the same disk pool as the base volume. Snapshot Volume A Snapshot volume is a logical entity that is structured on top of a Snapshot image, as illustrated in Figure 4. Figure 4) Snapshot volume. Snapshot images are not directly read/write accessible to hosts; to map a Snapshot image to a host, you must create a Snapshot volume from the Snapshot image. You can create up to four Snapshot volumes from the same Snapshot image. Every Snapshot volume has its own repository to track data changes that occur on the base volume. A one-to-one relationship exists between a Snapshot volume and its associated repository volume. The repository is not accessible for host I/O; it is used only internally to manage the Snapshot point-in-time image. The Snapshot volume has the same characteristics as the base volume (such as RAID level and I/O characteristics). However, the Snapshot volume is viewed as a separate standard volume by the storage system; therefore, it can be mapped to a host, read from, or written to. A Snapshot volume can be designated as read-only or read/write. If the Snapshot volume is read-only, then no associated repository is required. Read-only Snapshot volumes are generally designated for a backup application. Read/write Snapshot volumes are mapped to a host application for testing or experimental purposes. 33 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

34 A Snapshot volume can also be created from a consistency group, generating an exact point-in-time copy of all member volumes in that consistency group. Consistency Group A consistency group is a convenience entity that allows a collection of base volumes to have consistent behavior and interactions relative to Snapshot images and Snapshot volumes. Each volume is referred to as a member volume of the consistency group. A member volume is either a standard volume in a disk pool or volume group or a thin-provisioned volume in a disk pool. When a member volume is added to a consistency group, the system automatically creates an associated member repository and a Snapshot group for it. When a Snapshot image or Snapshot volume is created at the consistency group level, all associated member volumes in the consistency group are affected. A member volume in a consistency group receives an associated repository when it is added to the consistency group, as shown in Figure 5. If a Snapshot volume is created at the consistency group level, the member volume also receives another repository. A consistency group can have Snapshot images that are created manually or by applying a schedule. All Snapshot images in a consistency group have the same time stamp. Figure 5) Consistency group. The purpose of a consistency group is to make simultaneous Snapshot images of multiple volumes, which promotes consistent copies of a collection of volumes at a particular point in time. For example, you can create a consistency group to make a synchronized Snapshot image of several volumes that are in different volume groups or disk pools in a storage system. The consistency group mechanism is ideal for use with applications that span multiple volumes, such as a database application with the logs on one volume and the database on another volume. 34 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

35 Rollback It is sometimes necessary to return the data in a base volume to a previous point in time, possibly because of data corruption or user error on the base volume. This process is called a rollback from a previous Snapshot image. Snapshot rollback provides an instantaneous restore of a full volume from a Snapshot image, as shown in Figure 6. Figure 6) Rollback from Snapshot image to base volume. When you select a Snapshot image for rollback onto a base volume, all other Snapshot images (newer and older) in the Snapshot group are still preserved. A background process sweeps through the base volume s logical block accesses (LBAs) to find copy-on-write (COW) data in the rollback Snapshot image (or beyond) to be restored to the base volume. Rollback writes to the base volume might cause COW actions on the Snapshot group. Regardless of how much of the repository capacity is currently used, the base volume is accessible for both reads and writes during the rollback session, and all previously written data is also available immediately. Therefore, the associated repository must be large enough to contain all changes that might occur while the rollback operation is still processing. The data transfers from the repository back to the base volume continue as a background operation until the rollback is complete. Note: When a rollback has been initiated, you cannot stop or cancel it. I/O Handling to Snapshot Volumes The following sections describe the SANtricity COW Snapshot technology in general and then discuss what happens when Snapshot volumes receive ongoing write requests and read requests. 35 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

36 Snapshot COW Technology The Snapshot feature uses COW technology to create a durable point-in-time image that is useful for backup and testing. The Snapshot repository holds data that is saved from the base volume during COW activity, as illustrated in Figure 7. Figure 7) Snapshot COW technology. Original data blocks must first be written to the repository before they can be updated in the base volume, causing a slight performance impact while a Snapshot image or a Snapshot volume is enabled. Therefore, if a Snapshot image or a Snapshot volume is no longer needed, it should be disabled. If a Snapshot image is disabled, all COW activity from the base volume ceases, and the Snapshot volume becomes inaccessible for I/O operations, thus eliminating the performance impact. If the Snapshot volume is needed later, it can be recreated, which causes a new point-in-time image to be made, and the Snapshot volume becomes accessible for I/O operations again. Disabling the Snapshot volume is useful in situations that call for regular or periodic backups. The Snapshot volume can be disabled, but the configuration and structural elements (such as the repository, the Snapshot warning thresholds, and so on) are retained. The base volume can return to full performance for the period between backups. Then, at the time of the next backup, a new point-in-time image is created by using the previously defined Snapshot volume. When the repository of a Snapshot volume is nearing the full point, an alert is issued to the storage system. By default, this alert is issued when the repository is 75% full; however, as a storage administrator, you can adjust this threshold up or down as needed. When the alert is issued, you can increase the capacity of the repository at that time. In that regard, you can configure each Snapshot volume independently. Figure 8 depicts the COW technology. After the original data block has been copied to the repository, the block is overwritten on the base volume. The actual Snapshot image is created by combining the repository and the base volumes. 36 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

37 Figure 8) COW example. Write Requests After the Snapshot image is created, ensuing writes to the base volume cause COW actions to occur. With these COW actions, the required images of the affected sectors are saved in the Snapshot group repository before the base volume s sectors are overwritten with the new data. This process occurs only for the first overwrite of each base volume sector after the creation of the Snapshot image. In a Snapshot group, the creation of a new Snapshot image causes the following changes: The predecessor Snapshot image s content, including both its index and the captured data, is effectively frozen in the Snapshot group repository. That is, it is not changed thereafter, regardless of any ongoing write activity on the base volume. A new, and initially empty, index is created for the new Snapshot image. As base volume writes begin to occur, this new index is used when COW operations are carried out for the Snapshot group. Not only is the new index updated to track newly saved base volume data, but that new data is also added to the Snapshot group repository. In some cases, a given 64KB extent (called a cluster) of the base volume has been saved in the Snapshot group repository for a predecessor Snapshot image. In those cases, a rewrite of that cluster after the creation of the new Snapshot image causes another image of the cluster to be saved in the Snapshot group repository. As a result of these two points, the number of COW operations for a base volume write operation (when a Snapshot group exists for the base volume) is limited to at most one. Specifically, if the most recently created Snapshot index in the associated Snapshot group repository indicates that the affected cluster has already been saved in a previous COW for that Snapshot image, then no additional COW is required. Otherwise, a COW action is performed. This point holds true regardless of the number of predecessor Snapshot images that might exist. Only the current (newest) Snapshot image s index is used for determining whether a COW is needed when a base volume write request is processed. Read Requests With Snapshot groups, reading data associated with a Snapshot image starts out with the index for the relevant Snapshot image. That index is checked to determine whether the requested cluster is in that Snapshot image s portion of the Snapshot group repository. If so, the requested data is fetched from the Snapshot group repository. However, if the index indicates that the cluster is not present, the index of the next (that is, the successor, or later) Snapshot image in the Snapshot group is checked. This check is required because the base volume s cluster might have been written after the creation of the successor Snapshot image. In that case, the base volume data will have been saved to that successor Snapshot image s portion of the Snapshot group repository. However, the data that is saved there is actually the same content that must be returned when the predecessor Snapshot image is accessed. In fact, if a successor Snapshot image is found that does 37 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

38 contain the requested cluster, it is necessary to check the index of every successor in the Snapshot group. And they must be checked in the order in which they were created. If the requested cluster is not either in the Snapshot group repository of the requested Snapshot image or in any of its successors, then the data is retrieved from the base volume. It is retrieved from the base volume because clearly it has not been modified since the requested Snapshot image was created or since any of that Snapshot image s successors were created. When reading data from older Snapshot volumes, it is often necessary to check the indexes of many successor Snapshot images. The performance impact is mitigated by the extensive caching that occurs for Snapshot index information, so it is less likely that any actual media accesses are required to carry out the index checks. Table 4 summarizes how host I/O read and write requests are handled to a Snapshot volume. Table 4) I/O handling to a Snapshot base volume. I/O Type Reads to the base volume Writes to the base volume Reads to the Snapshot volume Writes to the Snapshot volume Handling A host I/O read to the base volume is passed through the base proxy with minimal interruption. When a write is destined for a base volume that has an associated Snapshot image, the write request first encounters the base proxy: If it is the first time that the base volume data has been modified, then the repository reads the appropriate LBAs from the base volume and writes them to the repository. If a Snapshot image already exists and it is the first time that the base volume data has changed since the last Snapshot image was created, then the repository reads the appropriate LBAs from the base volume and writes them to the Snapshot group repository. After the base proxy has been notified that this operation is complete, the host proceeds with the write operation in a normal fashion. If a second write request or iterative write requests occur to the same LBAs, but no new Snapshot images were made, no data movement is required to preserve the Snapshot image. Reads to the Snapshot volume occur as normal read requests, although they might be redirected to either the associated repository or the base volume, depending on the location of the most current data. If there are multiple Snapshot images of a given base volume in the same Snapshot group, then traversal of all the associated Snapshot image indexes might be required to locate the correct data. All writes to the Snapshot volume are passed to an associated user-created repository volume for that Snapshot volume. A corresponding pointer is also updated to reflect the position of this new data for future read requests to the Snapshot volume. Snapshot volumes might be mounted as read-only, in which case a specific Snapshot volume repository does not need to be used to store subsequent write modifications to the Snapshot volume itself. A single Snapshot group can contain up to 32 Snapshot images of a given base volume. Therefore, it is possible, albeit very unlikely, that the required capacity for the Snapshot group repository could be up to 32 times the capacity of the original base volume. For this scenario to occur, everything in the base volume would have to change between each subsequent Snapshot image being created. In addition to the COW clusters being copied to the Snapshot group repository, an additional small amount of metadata is associated with each Snapshot image. The metadata typically does not exceed 0.02% of the capacity of the original base volume and is therefore negligible. Accounting for Snapshot group repository capacity is largely a function of understanding the ratio of cluster overwrites as it relates to the base volume and the associated Snapshot images. 38 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

39 Appendix B: Manual Procedures for Image-Based Backup/Restore Previous versions of Cloud Connector required the user to manually perform steps to create and map the volumes used in performing image-based backups and restores. These manual steps are now automated when using the GUI. However, if you are using the Cloud Connector REST APIs directly, the manual steps might still be necessary. This appendix details the manual steps needed to perform image-based backup and restore. B.1 Performing an Image-Based Backup The following steps are required for an image-based backup. In this example, you back up all the data from UserData_Vol_1. Use these steps whether the backup target is AWS, NetApp StorageGRID, or NetApp Cloud Backup. Verify Backup Snapshot Volume Is Visible to SANtricity Cloud Connector Host Currently, it is the responsibility of the user to verify that the Snapshot backup volume is visible to the host through DM-Multipath. The quickest way to determine whether the Snapshot backup volume is visible is to create a new backup job with the backup summary screen. Select the image-based backup option, enter a name for the job, and click Next. 39 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

40 The Create Backup Job dialog box shows you what Snapshot volumes are available. If the Snapshot volume for UserData_Vol_1 appears, skip to the next section, Create a Backup Job. Otherwise, proceed to the next step. Launch NetApp SANtricity (or SANtricity Storage Manager) and define a new host to represent the SANtricity Cloud Connector host and associate the FC host ports. Verify that the operating system type is set to Linux (DM-MP). If you have not yet done so, create a backup Snapshot image of the base volume you want to back up. Note: Before creating the backup Snapshot image, verify that you have shut down or temporarily suspended any applications writing to the base volume. Map the appropriate backup Snapshot volume to the newly defined host. Note the worldwide names (WWNs) of the backup Snapshot volume mapped to the host. This information is used in subsequent steps to verify that the SANtricity Cloud Connector host has detected the volume correctly (Note: The WWNs you see on your setup differ from the WWNs shown here.) Log in to your SANtricity Cloud Connector host and verify the current status of DM-Multipath. [root@icta-faulk /] # multipath -ll [root@icta-faulk /] # 40 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

41 Instruct the SANtricity Cloud Connector host to rescan for newly attached devices. ~]# rescan-scsi-bus.sh -m -r Syncing file systems Scanning SCSI subsystem for new devices and remove devices that have disappeared Scanning host 0 for SCSI target IDs , all LUNs Scanning for device OLD: Host: scsi0 Channel: 02 Id: 00 Lun: 00 Vendor: DELL Model: PERC H310 Rev: 2.12 Type: Direct-Access ANSI SCSI revision: 05 Scanning host 1 for SCSI target IDs , all LUNs Scanning host 2 for all SCSI target IDs, all LUNs Scanning for device NEW: Host: scsi2 Channel: 00 Id: 00 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg18 changed: LU not available (PQual 1) REM: Host: scsi2 Channel: 00 Id: 00 Lun: 07 DEL: Vendor: NETAPP Model: Universal Xport Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg15 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 01 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 sg17 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 02 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 Scanning for device NEW: Host: scsi2 Channel: 00 Id: 07 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 additional output removed for brevity Attempting to update multipath devices... Removing multipath mappings for removed devices if all paths are now failed... Trying to discover new multipath mappings for newly discovered devices... 4 new or changed device(s) found. [2:0:0:0] [2:0:7:0] [10:0:0:0] [10:0:7:0] 0 remapped or resized device(s) found. 0 device(s) removed. Check the status of multipath again. [root@icta-faulk ~]# multipath -ll 3600a d7e400009eb75a1de1cd dm-2 NETAPP,INF size=500g features='4 queue_if_no_path pg_init_retries 50 retain_attached_hw_handle' hwhandler='1 rdac' wp=rw -+- policy='service-time 0' prio=14 status=active - 2:0:7:0 sdf 8:80 active ready running `- 10:0:7:0 sdc 8:32 active ready running `-+- policy='service-time 0' prio=9 status=enabled - 2:0:0:0 sde 8:64 active ready running `- 10:0:0:0 sdb 8:16 active ready running [root@icta-faulk ~]# As shown in this example, one new backup Snapshot volume has been discovered. The outlined information shows that the WWN of this volume matches the WWN information reported from SANtricity. Note: The leading three digits in the WWN can be ignored. Create a Backup Job To create a backup job, complete the following steps: 41 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

42 Select the backup Snapshot volume you want to back up from the Create Backup Job dialog box and click Finish. After you select the volume for backup and click Finish, you are asked whether you want to create another backup. Skip creation of another backup. The backup job summary screen shows you the status of all jobs. 42 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

43 Remove Backup Snapshot Volume from SANtricity Cloud Connector Host When the backup job has completed, you must manually remove the backup Snapshot volume from the SANtricity Cloud Connector host. Unmap the backup Snapshot volume from the SANtricity Cloud Connector host with SANtricity. Log back in to your SANtricity Cloud Connector host and initiate another rescan of the SCSI bus to remove the backup Snapshot volume. ~]# rescan-scsi-bus.sh -m -r Verify that multipath no longer detects the device. ~]# multipath -ll ~]# Delete the backup Snapshot volume and Snapshot image from the storage array. B.2 Performing an Image-Based Restore The following steps are required for an image-based restore. In this example, you restore all the data to R5_ImageRestore_Volume. Use these steps whether the restore source is AWS, NetApp StorageGRID, or NetApp Cloud Backup. Verify SANtricity Cloud Connector Host Has Visibility to the Restore Volume Currently, it is your responsibility to verify that the restore volume is visible to the host through DM- Multipath. The quickest way to determine if the restore volume is visible is to create a new restore job with the Restore Summary screen. Select the job from which you want to restore. 43 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

44 Next, you are asked to select the restore volume to use for the restore. If the restore volume (R5_ImageRestore_Volume) appears, skip to the next section, Create a Restore Job. Otherwise, proceed to the next step. Launch NetApp SANtricity System Manager or SANtricity Storage Manager (depending on E-Series model) and define a new host to represent the SANtricity Cloud Connector host and associate the FC host ports. Verify that the operating system host type in SANtricity Manager is set to Linux (DM- MP). Map the appropriate restore volume to the newly defined host. Note: The restore volume must be mapped only to the SANtricity Cloud Connector host for the restore operation. After the restore is complete, the volume can be mapped to another host. Note the WWNs of the restore volume mapped to the host. This information is used in subsequent steps to verify that the SANtricity Cloud Connector host has detected the volumes correctly. Figure 9) WWN of restore volume. Note: The WWNs you see on your setup differ from the WWNs shown here. Log in to your SANtricity Cloud Connector host and verify the current status of DM-Multipath. [root@icta-faulk /] # multipath -ll [root@icta-faulk /] # Instruct the SANtricity Cloud Connector host to rescan for newly attached devices. [root@icta-faulk ~]# rescan-scsi-bus.sh -m -r Syncing file systems Scanning SCSI subsystem for new devices and remove devices that have disappeared 44 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

45 Scanning host 0 for SCSI target IDs , all LUNs Scanning for device OLD: Host: scsi0 Channel: 02 Id: 00 Lun: 00 Vendor: DELL Model: PERC H310 Rev: 2.12 Type: Direct-Access ANSI SCSI revision: 05 Scanning host 1 for SCSI target IDs , all LUNs Scanning host 2 for all SCSI target IDs, all LUNs Scanning for device NEW: Host: scsi2 Channel: 00 Id: 00 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg18 changed: LU not available (PQual 1) REM: Host: scsi2 Channel: 00 Id: 00 Lun: 07 DEL: Vendor: NETAPP Model: Universal Xport Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg15 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 01 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 sg17 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 02 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 Scanning for device NEW: Host: scsi2 Channel: 00 Id: 07 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 additional output removed for brevity Attempting to update multipath devices... Removing multipath mappings for removed devices if all paths are now failed... Trying to discover new multipath mappings for newly discovered devices... 4 new or changed device(s) found. [2:0:0:0] [2:0:7:0] [10:0:0:0] [10:0:7:0] 0 remapped or resized device(s) found. 0 device(s) removed. Check the status of multipath again. [root@icta-faulk ~]# multipath -ll 3600a d ae75a1ddcc1 dm-3 NETAPP,INF size=500g features='4 queue_if_no_path pg_init_retries 50 retain_attached_hw_handle' hwhandler='1 rdac' wp=rw -+- policy='service-time 0' prio=14 status=active - 2:0:7:0 sdf 8:80 active ready running `- 10:0:7:0 sdc 8:32 active ready running `-+- policy='service-time 0' prio=9 status=enabled - 2:0:0:0 sde 8:64 active ready running `- 10:0:0:0 sdb 8:16 active ready running [root@icta-faulk ~]# As shown in this example, one new volume has been discovered. The outlined information shows that the WWN of this volume matches the WWN information reported from SANtricity. The leading three digits in the WWN can be ignored. Create a Restore Job The restore job summary screen shows a list of backup jobs from which you can restore. Use the backup job created in the section Performing an Image-Based Backup. 45 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

46 Select the restore volume and then click Finish. 46 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

47 The restore job summary screen now shows the status of the restore job you just scheduled. Remove the Restore Volume from the SANtricity Cloud Connector Host When the restore job has completed, you are responsible for removing the restore volume from the SANtricity Cloud Connector host. Unmap the restore volume from the SANtricity Cloud Connector host with SANtricity. Log back in to your SANtricity Cloud Connector host and initiate another rescan of the SCSI bus to remove the restore volume. ~]# rescan-scsi-bus.sh -m -r Verify that multipath no longer detects the device. ~]# multipath -ll ~]# Appendix C: Manual Procedures for File-Based Backup and Restore Previous versions of Cloud Connector required the user to manually perform steps to create and map the volumes used in performing image-based backups and restores. These manual steps are now automated when using the GUI, but if you are using the Cloud Connector REST APIs directly, the manual steps might still be necessary. This appendix details the manual steps needed to perform file-based backup and restore. C.1 Performing File-Based Restore The following steps are required for a file-based backup. In this example, UserData_Vol_1 contains a Linux ext4 file system on it. Use these steps whether the backup target is AWS, NetApp StorageGRID, or NetApp Cloud Backup. 47 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

48 Note: A file-based backup unconditionally backs up all files on the file system you specify. However, you can perform a selective restore of files and folders. Verify SANtricity Cloud Connector Host Has Visibility to the File System To perform a file-based backup: SANtricity Cloud Connector must have visibility to the E-Series array that contains the Snapshot backup volume, which is visible through DM-Multipath. The Snapshot backup volume must have the file system you want to back up. The file system must be mounted to the SANtricity Cloud Connector host. The quickest way to determine if the Snapshot backup volume and file system are visible is to complete the following procedure: Create a backup job with the Backup summary screen. Select the folder/file-based option, enter a job name, and click Next. The Create Backup Job dialog box shows you what file systems are available. If the backup Snapshot volume has been mounted, it appears in this dialog box, and you can skip to the section Create a Backup Job. Otherwise, proceed to the next step. 48 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

49 Launch NetApp SANtricity System Manager (or SANtricity Storage Manager for older generation E- Series arrays) and define a new host to represent the SANtricity Cloud Connector host and associate the FC host ports. Verify that the operating system type is set to Linux (DM-MP). If you have not yet done so, create a backup Snapshot image of the base volume that contains the file system you want to back up. Note: Before creating the backup Snapshot image, verify that you have shut down or temporarily suspended any applications writing to the base volume. Map the appropriate backup Snapshot volume to the newly defined host. Note the WWNs of the backup Snapshot volume mapped to the host. This information is used in subsequent steps to verify that the SANtricity Cloud Connector host has detected the Snapshot volume correctly. (Note: The WWNs you see on your setup differ from the WWNs shown here.) Figure 10) WWN of Snapshot volume. Log in to your SANtricity Cloud Connector host and verify the current status of DM-Multipath. [root@icta-faulk /] # multipath -ll [root@icta-faulk /] # Instruct the SANtricity Cloud Connector host to rescan for newly attached devices. [root@icta-faulk ~]# rescan-scsi-bus.sh -m -r Syncing file systems Scanning SCSI subsystem for new devices and remove devices that have disappeared Scanning host 0 for SCSI target IDs , all LUNs Scanning for device OLD: Host: scsi0 Channel: 02 Id: 00 Lun: 00 Vendor: DELL Model: PERC H310 Rev: NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

50 Type: Direct-Access ANSI SCSI revision: 05 Scanning host 1 for SCSI target IDs , all LUNs Scanning host 2 for all SCSI target IDs, all LUNs Scanning for device NEW: Host: scsi2 Channel: 00 Id: 00 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg18 changed: LU not available (PQual 1) REM: Host: scsi2 Channel: 00 Id: 00 Lun: 07 DEL: Vendor: NETAPP Model: Universal Xport Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg15 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 01 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 sg17 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 02 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 Scanning for device NEW: Host: scsi2 Channel: 00 Id: 07 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 additional output removed for brevity Attempting to update multipath devices... Removing multipath mappings for removed devices if all paths are now failed... Trying to discover new multipath mappings for newly discovered devices... 4 new or changed device(s) found. [2:0:0:0] [2:0:7:0] [10:0:0:0] [10:0:7:0] 0 remapped or resized device(s) found. 0 device(s) removed. Check the status of multipath again. [root@icta-faulk ~]# multipath -ll 3600a d ab35a156d7e dm-2 NETAPP,INF size=500g features='4 queue_if_no_path pg_init_retries 50 retain_attached_hw_handle' hwhandler='1 rdac' wp=rw -+- policy='service-time 0' prio=14 status=active - 2:0:7:0 sdf 8:80 active ready running `- 10:0:7:0 sdc 8:32 active ready running `-+- policy='service-time 0' prio=9 status=enabled - 2:0:0:0 sde 8:64 active ready running `- 10:0:0:0 sdb 8:16 active ready running [root@icta-faulk ~]# As shown in this example, one new volume has been discovered. The outlined information shows that the WWN of this volume matches the WWN information reported from SANtricity System Manager or Storage Manager. The leading three digits in the WWN can be ignored. Mount the file system located on this Snapshot backup volume. [root@icta-faulk ~]# mount /dev/dm-2 /mnt/scc [root@icta-faulk ~]# Create a Backup Job In the Create Backup Job dialog box, select the file system from which you want to back up. Then click Finish. 50 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

51 After you select the file system for backup and click Finish, you are asked whether you want to create another backup. Skip creation of another backup. The Backup job summary in the following image shows you the status of all jobs. Remove Backup Snapshot Volume from SANtricity Cloud Connector Host When the backup job has completed, you are responsible for removing the restore volume from the SCC host. Unmount the file system that was mounted to create the backup. 51 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

52 ~]# umount /mnt/scc Unmap the Snapshot backup volume from the SANtricity Cloud Connector host using SANtricity System Manager or Storage Manager. Log back in to your SANtricity Cloud Connector host and initiate another rescan of the SCSI bus to remove the Snapshot backup volume. ~]# rescan-scsi-bus.sh -m -r Verify that multipath no longer detects the device. ~]# multipath -ll ~]# Delete the Snapshot backup volume and Snapshot image from the storage array. C.2 Performing File-Based Restore The following steps are required for a file-based restore. In this example, you restore the contents of several folders to a partition that has been created on a restore volume (RestoreData_Vol_1). The partition is mounted as /mnt/scc. Use these steps whether the restore source is AWS, NetApp StorageGRID, or NetApp Cloud Backup. Verify SANtricity Cloud Connector Host Has Visibility to the Restore Volume To perform a file-based restore: NetApp SANtricity Cloud Connector must have visibility to the E-Series array that contains the restore volume, which is visible through DM-Multipath. SANtricity Cloud Connector must have visibility to an ext4 Linux file system partition that has already been created on the restore volume. The file system must be mounted to the SANtricity Cloud Connector host. The quickest way to determine if the restore volume is visible is to create a file-based restore job with the restore summary screen. Select the previously created file-based backup job from which you want to restore and then click Next. 52 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

53 Next, you are asked to specify the files you want restored. Because you don t know if the volume you want to restore is visible, skip through this screen by clicking Next. 53 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

54 Next, you are asked to select the restore volume to use for the restore. If the partition associated with the restore volume has been mounted, it appears in this dialog box, and you can skip to the next section, Create a Restore Job. Otherwise, proceed to the next step. 54 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

55 Launch SANtricity (or SANtricity Storage Manager) and define a new host to represent the SANtricity Cloud Connector host and associate the FC host ports. Verify that the operating system type is set to Linux (DM-MP). Map the appropriate restore volume to the newly defined host. Note: The restore volume must be mapped only to the SANtricity Cloud Connector host for the restore operation. After the restore is complete, the volume can be mapped to another host. Note the WWNs of the restore volume mapped to the host. This information is used in subsequent steps to verify that the SANtricity Cloud Connector host has detected the volumes correctly. Note: The WWNs you see on your setup differ from the WWNs shown here. Log in to your SANtricity Cloud Connector host and verify the current status of DM-Multipath. [root@icta-faulk /] # multipath -ll [root@icta-faulk /] # Instruct the SANtricity Cloud Connector host to rescan for newly attached devices. [root@icta-faulk ~]# rescan-scsi-bus.sh -m -r Syncing file systems Scanning SCSI subsystem for new devices and remove devices that have disappeared Scanning host 0 for SCSI target IDs , all LUNs Scanning for device OLD: Host: scsi0 Channel: 02 Id: 00 Lun: 00 Vendor: DELL Model: PERC H310 Rev: 2.12 Type: Direct-Access ANSI SCSI revision: 05 Scanning host 1 for SCSI target IDs , all LUNs Scanning host 2 for all SCSI target IDs, all LUNs Scanning for device NEW: Host: scsi2 Channel: 00 Id: 00 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg18 changed: LU not available (PQual 1) REM: Host: scsi2 Channel: 00 Id: 00 Lun: 07 DEL: Vendor: NETAPP Model: Universal Xport Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 sg15 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 01 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 sg17 changed: LU not available (PQual 1) OLD: Host: scsi2 Channel: 00 Id: 02 Lun: 00 Vendor: NETAPP Model: INF Rev: 0840 Type: Direct-Access ANSI SCSI revision: 05 Scanning for device NEW: Host: scsi2 Channel: 00 Id: 07 Lun: 00 Vendor: NETAPP Model: INF Rev: 0825 Type: Direct-Access ANSI SCSI revision: 05 additional output removed for brevity Attempting to update multipath devices... Removing multipath mappings for removed devices if all paths are now failed... Trying to discover new multipath mappings for newly discovered devices NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

56 4 new or changed device(s) found. [2:0:0:0] [2:0:7:0] [10:0:0:0] [10:0:7:0] 0 remapped or resized device(s) found. 0 device(s) removed. Check the status of multipath again. [root@icta-faulk ~]# multipath -ll e c559b67bef dm-3 NETAPP,INF size=500g features='4 queue_if_no_path pg_init_retries 50 retain_attached_hw_handle' hwhandler='1 rdac' wp=rw -+- policy='service-time 0' prio=14 status=active - 2:0:7:0 sdf 8:80 active ready running `- 10:0:7:0 sdc 8:32 active ready running `-+- policy='service-time 0' prio=9 status=enabled - 2:0:0:0 sde 8:64 active ready running `- 10:0:0:0 sdb 8:16 active ready running [root@icta-faulk ~]# As shown in this example, one new volume has been discovered. The outlined information shows that the WWN of this volume matches the WWN information reported from SANtricity. (The leading three digits in the WWN can be ignored.) Mount the file system present on the restore volume. [root@icta-faulk ~]# mount /dev/dm-3 /mnt/scc [root@icta-faulk ~]# Create a Restore Job The Restore job summary screen shows a list of backup jobs from which you can restore. Select the backup job created in the section Performing File-Based Backup. 56 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

57 Next, you are asked to specify the files that you want restored. In this example, the Project 1 and Project 2 folders have been selected. All files and subfolders in these folders are restored. Click Next to proceed to the next screen. 57 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

58 Select the restore volume from the list that appears. 58 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

59 The restore job summary screen now shows the status of the restore job you just scheduled. Remove the Restore Volume from the SCC Host When the restore job has completed, you are responsible for removing the restore volume from the SANtricity Cloud Connector host. Unmount the file system that was mounted for the restore. ~]# umount /mnt/scc Unmap the restore volume from the SCC host with SANtricity. Log back in to your SCC host and initiate another rescan of the SCSI bus to remove the restore volume. ~]# rescan-scsi-bus.sh -m -r Verify that multipath no longer detects the device. ~]# multipath -ll ~]# 59 NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

60 Where to Find Additional Information The following references were used in this TR: E-Series Snapshot Quick Start Guide Deploying NetApp E-Series and EF-Series Copy Services with Oracle and SQL Server Databases SANtricity Cloud Connector SANtricity Web Services Proxy Version History Version Date Document Version History Version 1.0 January 2018 Initial release. Version 2.0 June 2018 Revised for version NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.

61 Refer to the Interoperability Matrix Tool (IMT) on the NetApp Support site to validate that the exact product and feature versions described in this document are supported for your specific environment. The NetApp IMT defines the product components and versions that can be used to construct configurations that are supported by NetApp. Specific results depend on each customer s installation in accordance with published specifications. Copyright Information Copyright 2018 NetApp, Inc. All rights reserved. Printed in the U.S. No part of this document covered by copyright may be reproduced in any form or by any means graphic, electronic, or mechanical, including photocopying, recording, taping, or storage in an electronic retrieval system without prior written permission of the copyright owner. Software derived from copyrighted NetApp material is subject to the following license and disclaimer: THIS SOFTWARE IS PROVIDED BY NETAPP AS IS AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE, WHICH ARE HEREBY DISCLAIMED. IN NO EVENT SHALL NETAPP BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. NetApp reserves the right to change any products described herein at any time, and without notice. NetApp assumes no responsibility or liability arising from the use of products described herein, except as expressly agreed to in writing by NetApp. The use or purchase of this product does not convey a license under any patent rights, trademark rights, or any other intellectual property rights of NetApp. The product described in this manual may be protected by one or more U.S. patents, foreign patents, or pending applications. RESTRICTED RIGHTS LEGEND: Use, duplication, or disclosure by the government is subject to restrictions as set forth in subparagraph (c)(1)(ii) of the Rights in Technical Data and Computer Software clause at DFARS (October 1988) and FAR (June 1987). Trademark Information NETAPP, the NETAPP logo, and the marks listed at are trademarks of NetApp, Inc. Other company and product names may be trademarks of their respective owners. TR NetApp SANtricity Cloud Connector NetApp, Inc. All rights reserved.