FAQ

This section answers typical questions from customers concerning NEC Storage HS Series.

General

Q.1What is NEC Storage HS?

A.1NEC Storage HS is a grid-based secondary storage solution unlike traditional tape or disk-based systems. It utilizes data compression (DataRedux), automatic optimization (DynamicTopology), and distributed data redundancy (Distributed Resilient Data) technologies.

Q.2What advantages does NEC Storage HS have over traditional tape library backup systems?

A.2Being a disk-based solution, the labor involved in handling and managing tape cartridges as well as the physical space required to store them is no longer necessary. Also, with the help of data reduction technologies, costs per GB comparable to tape can be realized.

Q.3What kind of remote backup/disaster recovery features are available?

A.3

Filesystems can be duplicated across systems using NEC Storage HS's replication feature (RepliGrid).

Q.4

What backup servers/backup software can be used with NEC Storage HS?

A.4Refer to the brochure of NEC Storage HS.

Q.5Does NEC Storage HS come with a custom cabinet (rack)?

A.5A custom cabinet is not included.

Q.6Is it possible to expand the system without stopping operation?

A.6

The scale-out model can be expanded without stopping operation, due to its automatic optimization technology (DynamicTopology).

Only in the following cases, the system will need to be stopped while upgrading:

- When adding a Hybrid node (HN) or a Storage Node (SN) to a one HN configuration.

- When adding a Hybrid node (HN) to a two-or-more node configuration containing only one HN.

- When adding L2 switch.

- When installing an upgrade kit.

 

The single node model cannot be expanded witout stopping operation because it's necessary to install an upgrade kit in order to expand the capacity.

Q.7

Does NEC Storage HS support Active Directory (*1)?

(*1) A directory service developed by Microsoft

A.7Yes, Active Directory is supported.

Q.8What are the licensing requirements when using replication?

A.8

Scale-out model:

Licenses are required for both the Hybrid node where the master file system resides, and the Hybrid node on the replica system, where the replica file system resides. Replication can still be configured on Hybrid nodes without a license; however, replication will fail on execution.

If failover occurs while using AN clustering, licenses are carried over to the failover target node.

 

Single node model:

Licenses are required for both the master and the replica system. Replication can still be configured on systems without a license; however, replication will fail on execution.

Q.9Is NEC Storage HS recognized as a virtual tape library by the backup server, or as a disk?

A.9NEC Storage HS is recognized as a disk (network share).

Q.10 What are Hybrid nodes (HN) and Storage nodes (SN)?

A.10

 Scale-out NEC Storage HS systems are made up of servers called Hybrid nodes (HN) and Storage nodes (SN), each having a different role.

HN: Control external I./O, manage filesystems, and provide storage space. Increasing the number of HNs in a NEC Storage HS system will increase its performance and capacity.

SN: Provide storage space. Increasing the number of SNs in a NEC Storage HS system will increase its capacity.

 

Single node model consists of only one type of node.

Q.11

 Does NEC Storage HS support Access Control Lists (ACL *1)?

 *1) A method of managing the access permissions of files and folders on Windows operating systems

A.11

 Yes, available to use ACL. In order to control access by ACL fearure, "Active Directory" should be selected as the CIFS authentification method. However, the ACL of local users that are not managed by Active Directory cannot be specified to the file on NEC Storage HS.

 

The auditing feature of ACL is not available on NEC Storage HS.

Features/Performance

Q.1What does the data compression (DataRedux) feature do?

A.1Data compression (DataRedux) enables the highly efficient use of storage space by storing only compressed, unique data. Data to be written to a disk is checked against existing data. If duplicate data is found (even if in different files), this data is prevented from being stored twice, with the existing data being shared instead between files.

Q.2What kind of benefits does the data compression (DataRedux) feature provide?

A.2

Data compression (DataRedux) provides the following benefits:

- As only compressed, unique data is stored, processing time and I/O load is reduced.

- When performing generational backups, the large amount of duplicate data present results in both backup times and disk usage being greatly reduced.

- As the data compression (DataRedux) feature is also utilized by replication, only compressed data that does not already exist at the replica site needs to be transferred. This enables remote replication to be performed even on low speed connections.

Q.3What does the automatic optimization (DynamicTopology) feature do?

A.3

Automatic optimization (DynamicTopology) is a feature available on the scale-out model that allows nodes to be added without stopping the system, enabling performance and capacity to be increased dynamically on the fly. (This feature is not available on the single-node model).

After adding a new node, existing data is automatically redistributed across the system for optimal performance and data resiliency, which simplifies management.

Q.4What does the distributed data redundancy (Distributed Resilient Data) feature do?

A.4

Distributed data redundancy (Distributed Resilient Data) breaks backup data into multiple blocks and adds special redundancy codes before storing it in a distributed fashion across several disks. This allows the original data to be reconstructed even after failures at multiple storage locations.

The single-node model is capable of withstanding up to three simultaneous disk drive failures. (*1).

The scale-out model is able to withstand up to six simultaneous disk drive failures not only within a single node, but also spread over multiple nodes (*2).

Offering greater reliability and higher simultaneous fault tolerance than RAID, in the event of a failure, Distributed Resilient Data prevents the loss of data.

 

  (*1) With the default number of redundancy codes (parities).

  (*2) Depending on the node configuration and number of redundancy codes (parities).

          With the default number of parities, NEC Storage HS can withstand three simultaneous

          failures.

          By increasing the number of parities, this can be increased to up to six simultaneous

          failures.

Q.5Can information about NEC Storage HS's storage capacity and other statisics be output to a file?

A.5Performance and capacity statistics can be output to a file in Comma-Seperated Value (CSV) format from the Graphical User Interface (GUI).

Q.6What is the space reclamation feature?

A.6Space reclamation refers to the process of recovering free space that has been made available after deleting files. It is possible that due to the data compression (DataRedux) feature, data from the deleted file may be shared by other files. In this case, the actual data will not be deleted until all files that share it have been deleted.

Q.7

Can the following tasks be performed while space reclamation is running?

  - Backup

  - Writing data

  - Reading data

A.7

Space reclamation can run in one of two different modes - Foreground (FG) or Background (BG) mode. Which one of these modes space reclamation runs in can be set from the Command Line Interface (CLI) (*1). The tasks that can be performed while space reclamation is executing vary depending on whether reclamation runs in the FG or BG mode.

 

- When running in BG mode

  All tasks (backup, data read/write) can be performed.

- When running in FG mode

  Data can be read, but not written. Backup can not be performed.

 

(*1) The default setting for a new NEC Storage HS installation is BG mode. When upgrading, the existing space reclamation mode setting is preserved.

Q.8What is marker filtering?

A.8

Some commercially available backup software products insert 'markers' (management information) into backup data when performing backups.

These markers can have an adverse affect on compression, lowering the deduplication ratio of data stored in the system.

The optional marker filtering feature works with supported backup software products to filter markers existing in backup data during a backup, thereby increasing data deduplication performance.

Q.9Will marker filtering work on all backup software?

A.9

No, the backup software to target is specified for each filesystem based on the marker type set at filesystem creation.

For details on verified backup software, refer to the brochure of NEC Storage HS.

Q.10 Will doubling the number of Hybrid nodes (HN) result in a doubling of performance?

A.10 For the scale-out model, increasing the number of HNs will result in a near-linear increase in performance. However, if the number of Storage nodes (SN) is less than the number of HNs, whether or not performance will scale linearly is dependant on the system configuration.

Q.11 Can the system continue operating when a node failure occurs?

A.11

 Scale-out model

Even if a node fails, as long as a healthy Hybrid node (HN) exists and data can be restored from the remaining nodes, the system will be able to continue operating.

 

- Existence of a healthy HN

In configurations containing two or more HNs, if an HN fails and the AN clustering feature is enabled, failover (refer to Q2-13) will occur, causing the failed HNs services to be taken over by another HN. However, access will temporarily be unavailable after failover when connecting via CIFS.

 

- Recovering data from remaining healthy nodes

In the following configrations, data will be able to be restored after a single node failure.

  - when number of parities is one : 12 or more nodes

  - when number of parities is two : 6 or more nodes

  - when number of parities is three : 4 or more nodes

  - when number of parities is four : 3 or more nodes

  - when number of parities is five : 3 or more nodes

  - when number of parities is six : 2 or more nodes

To configure the above parity number is strongly recommended.

 

Single-node model

After a node failure, operation can not be continued.

Q.12 When a disk failure occurs, what steps need to be performed to restore the data?

A.12 No action needs to be taken. Lost data will be automatically reconstructed and restored onto the remaining disks. Additionally, on replacement of the failed disk, data will automatically be redistributed over all disks (including the newly replaced disk).

Q.13

 What is failover?

What conditions will cause a failover to occur?

A.13

 Failover refers to a feature that allows services provided by the accelerator function nodes that cannot continue operating (due to a stall or crash) to be taken over by another, healthy Hybrid node (HN). This feature can be used in configurations with more than one HN.

 

Failover can be caused by events such as a stalled OS, reboot, shutdown, or a power outage. An external network failure can also be set as a condition that will trigger a failover.

Q.14 What kinds of notifications can be sent when a failure occurs?

A.14

 SNMP or mail notifications.

SNMP versions V1 and V2c are supported.

Mail notifications support SMTP authentication. The mail address that notifications will be sent to can be set for each failure severity level.

Q.15 What happens if a disk failure occurs during backup/restore?

A.15

 Disk failures during backup/restore can cause currently executing backup/restore jobs to fail.

Certain client configuration settings (*1) can be made so that backup/restore jobs will not be affected by a single disk failure.

Even if the above configuration settings have been made, if two disks fail within 30 minutes of each other, multiple disk failure is considered to have occurred and backup/restore jobs may fail. Even in this case, no data will be lost.

 

(*1) If the filesystem backup/restore export type is NFS

        When NFS mounting the NEC Storage HS filesystem use the 'hard' option.

 

        If the filesystem backup/restore export type is CIFS

        On the Windows backup server, set the value of session timeout to at least 10 minutes.

Q.16 Can filesystems be optimized for different purposes (examples: fault-tolerance and storage efficiency)?

A.16

 Yes, you can choose to optimize for fault-tolerance or storage efficiency.

This is accomplished by setting the number of redundancy codes (parities), which can be configured per filesystem, to an appropriate value.

Increasing the number of parities improves the fault-tolerance of the filesystem while decreasing the number of parities increases storage efficiency.

Q.17 Are there any notes that should be concerned when storing file in a directory?

A.17 NEC Storage HS takes longer time to display the list of files compared to general storages. It will take especially long when the system is overloaded. Therefore, storing many files in one directory is not recommended.

Q.18 What does the data tampering protection (WORM) feature do?

A.18

 Write Once Read Many (WORM) is a feature that protects files for a specified period from being overwritten, deleted, or otherwise modified either by accident or through tampering. Protected files will not be able to be modified in any way.

The protection period can be set by the backup application or by the user, either per filesystem, or per file.

NEC Storage HS supports WORM.

Q.19 When using data tampering protection (WORM), can a file protected with a protection period be modified once its protection period expires?

A.19 The file will not be able to be modified; however, it will be able to be deleted.

Q.20 When using data tampering protection (WORM), can a filesystem be deleted if it contains files with a protection period before the protection period is over?

A.20

 In order to use WORM, it must be enabled when creating a filesystem. The type of WORM protection, Compliance or Enterprise can also be selected.

Whether or not the filesystem can be deleted depends on the type of WORM protection selected.

A filesystem with Enterprise type WORM protection can be deleted by an administrator from the Command Line Interface (CLI), even if it contains protected files still in their protection period.

A filesystem with Compliance type WORM protection cannot be deleted if it contains protected files still in their protection period.

Q.21 When using data tampering protection (WORM), what happens if the default protection period for a file exceeds the latest possible protection period expiration date (2100/12/31)?

A.21 The file protection period expiration date becomes 2100/12/31. For example, if 100 years is specified for the filesystem default protection period, the protection period expiration date of the backed up files will become 2100/12/31.

Q.22 When using data tampering protection (WORM), what happens when a file protection period is set to a value shorter than the minimum protection period for the filesystem, or longer than the maximum protection period for the filesystem?

A.22 When a file protection period is set to a value shorter than the minimum protection period for the filesystem, the file protection period will be changed to the minimum protection period for the filesystem. When a file protection period is set to a value longer than the maximum protection period for the filesystem, the file protection period will be changed to the maximum protection period for the filesystem.

Q.23 What are the licensing requirements for using the data tampering protection (WORM) feature?

A.23

 Scale-out model

A license is required for each Hybrid node using WORM.

When using AN clustering, licenses are automatically carried over to the failover target on failover.

 

Single-node model

A license is required for each system using WORM.

Q.24 Can filesystems using data tampering protection (WORM) be replicated?

A.24 Yes, they can. However, NEC Storage HS WORM software licenses are required at both the replica and the master sites.

Q.25 NEC Storage HS allows the level of fault-tolerance of a filesystem to be set by configuring the number of redundancy codes (parities). Can replication be performed between a master and replica filesystem that have different fault-tolerance levels?

A.25 Yes, replication can be performed between a master and replica filesystem that have different fault-tolerance levels.

Q.26 Is it possible to place a quota on the amount of storage capacity users can use?

A.26

 Limits on capacity cannot be placed on users or user groups.

However, by setting a hard quota (the maximum amount of storage space that can be used by a filesystem) at filesystem creation time, a limit can effectively be placed on the amount of data that can be written by each filesystem (*1).

In some cases backup software itself can also place limits on storage capacity usage. If this is the case, the effective limit will be whichever is smaller, NEC Storage HS's hard quota or the backup software's limit.

 

(*1) This value set here refers to the amount of data written logically and is not the amount of data that is actually written to NEC Storage HS's hard disks.

Q.27 What happens if the size of a master filesystem exceeds the hard quota (the maximum amount of storage space that can be used by a filesystem) of the replica filesystem?

A.27 The replication job will fail. To avoid this, delete unneeded files from the master filesystem to reduce its size to less than the hard quota of the replica filesystem before performing replication.

Q.28 Is a replication job can start when the number of executing replication job is already maximum?

A.28

 Scheduled replication jobs cannot start when the number of executing replication job is already maximum. The scheduled replication jobs will start automatically if an executing job ends before ""Max Duration"" of the scheduled job. Otherwise it will not start.

 

The replication job executed by GUI or CLI will fail immediately and not restarted.

Q.29 What is the maximum number of simultaneous replication jobs?

A.29

 HS3-40 : 32

HS8-40 : 125 (per Hybrid Node)

Q.30 Are there any ways to improve the security of data transfers over the network during replication?

A.30 The security of data sent over the network during replication can be improved by using the replication data packet encryption feature. This feature can be enabled either when registering the remote system, or when changing its settings. The encryption algorithm used is Advanced Encryption Standard (AES).

Q.31 What web browsers does the NEC Storage HS Graphical User Interface (GUI) support?

A.31

 The NEC Storage HS GUI supports the following browsers:

- Internet Explorer : 8.0/9.0/10.0

Q.32 Are there any notes about remote execution of CLI command?

A.32

 When performing failover or takeback manually during remote execution of CLI command using rsh protocol, the result of commend execution cannot be returned. Use ssh protocol if getting result is necessary.

When failover or takeback occurs automatically due to a node failure or other reasons, the result of CLI command will not be returned even when using rsh protocol.

Configuration/Connection

Q.1Can redundant paths between the backup server and NEC Storage HS be setup?

A.1

Path redundancies can be setup by using port bonding. The types of bonding that can be used with NEC Storage HS's external network are listed below:

 

Networking standard 10Gbps

・balance-alb

・802.3ad

・active-backup

 

Networking standard not 10Gbps

・balance-rr

・balance-alb

・802.3ad

・active-backup

Q.2

Can NEC Storage HS be connected to a tape system and back up data to tape?

A.2

NEC Storage HS cannot backup to tape directly. In order to backup to tape, use a separate backup server connected to a tape system.

Q.3Is it possible to use NEC Storage HS with existing 10/100/1000 base hub?

A.3

NEC Storage HS supports 10/100/1000 base hubs. Depending on the specific hub used as well as port bonding settings, some hubs may be incompatible.

Q.4

Can monitors or input devices (example: a keyboard) be connected directly to the system?

A.4Monitors and input devices cannot be directly connected to NEC Storage HS. NEC Storage HS can be managed over the network from a client terminal.

Q.5What is the way of changing IP address of NTP server synchronized with NEC Storage HS?

A.5

When changing IP Address of NTP server synchronized with NEC Storage HS, change the setting of NTP server at first, then change the setting of NEC Storage HS as following steps :

 

[NTP server synchronized with NEC Storage HS]

 - Change IP address following the process for its OS.

 - Restart the NTP server

 

[NEC Storage HS]

 - Click "Others" on the Main Menu of GUI, set new IP address of NTP server in the Date & Time window.

 - Restart NEC Storage HS following the guidance on the screen.

Q.6

Connection from Windows Server 2008 R2 to NEC Storage HS fails.

- Error code: 0x80070035 The network path was not found

A.6This case corresponds to "kb2194664" of Microsoft Knowleadge Base. Install Service Pack 1 or later vertion to Windows Server 2008 R2.

Top of this page