Ex a Grid Admin Guide

ExaGrid Administrator’s Guide

PN: 210-0165-09

Copyright No part of this document may be reproduced or transmitted in any form or by any means without the express written permission of ExaGrid Systems, Inc. © 2003-2011 ExaGrid Systems, Systems, Inc. All rights reserved. Printed in USA. This document, the associated software, and the associated online documentation are the property of ExaGrid Systems, Inc. or its licensors, as applicable, and are loaned to the user under the terms of the ExaGrid Systems, Inc. End User License Agreement. Unauthorized copying or use of the software or any associated materials is contrary to the property rights of ExaGrid Systems and is a violation of state and federal law.

Trademarks ExaGrid is a registered trademark of ExaGrid Systems, Inc. ExaGrid Systems (Logo), InstantDR, InfiniteFiler, GRIDdisk and Intelligent Disk-based Data Protection are all trademarks of ExaGrid Systems, Inc. All third-party trademarks are the property of their respective owners.

License Agreement This document, the associated software and the associated online documentation documentation are the property of ExaGrid Systems, Inc. or its licensors, as applicable. The use of these materials and the software is strictly limited to those users who have signed the ExaGrid Systems, Inc. End User License Agreement. For any software acquired directly or indirectly on behalf of a unit or agency of the United States Government, Government, whether for civilian agencies or for units of the Department of Defense, the software is a commercial item or commercial computer software (and documentation), and pursuant to Section 12.212 of the Federal Acquisition Regulations (FAR) or DoD FAR Supplement Section 227.7202, is provided under restricted rights as enumerated in the End User License Agreement.

Contents

1.

The ExaGrid System

1

ExaGrid Server Configurations Configurations 2 Power Requirements 2 Environmental Requirements 3 ExaGrid Hub Sites and Spoke Sites 3 Data Flow 4 2.

NetWorking

6

Network Requirements 6 3.

Power Conditioning

9

4.

ExaGrid System Overview

10

Single-Site ExaGrid Systems 10 Multi-site ExaGrid Systems 11 5.

Shares

12

Share Protection 16 Share Migration 16 6.

The ExaGrid User Interface

Logging In

17

18

ExaGrid Administrator ’s Guide iii

7.

Creating Application Specific and Application Independent Shares

8.

Creating TempDR Shares

9.

ExaGrid Reports

22

23

Summary Reports

24

Replication Reports 27 Deduplication Reports 30 10.

Monitoring Your ExaGrid System

SMTP Email Notifications SNMP Traps

32

32

33

Status Monitor

33

ExaGrid MIB 34 11.

Restoring Files

39

12.

ExaGrid System Maintenance

40

Shutting Down and Restarting an ExaGrid Server Account Management

41

Adding Additional ExaGrid Servers 43 Adding an ExaGrid Site to an ExaGrid System 44 Record Keeping 46 13.

ExaGrid System Disaster Recovery

14.

Troubleshooting

Disk Failure

48

48

Power Supply Failure

51

Network Debugging Tools Notifications 53 Consumption 53 iv

ExaGrid Systems

52

47

41

19

Replication 15.

54

Performance Tuning

56

ExaGrid Administrator’s Guide v

vi

ExaGrid Systems

Preface This guide provides an overview of the ExaGrid System and getting started instructions for the novice ExaGrid Administrator. Detailed descriptions of the ExaGrid System and advanced administrator procedures can be found in the ExaGrid System’s online help.

Related Documentation Title

Part Number

Assembling and Initializing Your ExaGrid System

210-0164-xx

ExaGrid Release Notes

210-0174-xx

Contacting ExaGrid Technical Support If you have technical questions about this product that are not answered in this document contact ExaGrid Support: ■ ■

Email: [email protected] Phone: 800-868-6985 and at the prompt, press 2.


vii

Glossary Understanding these terms will help you better understand the ExaGrid System. Byte-level data deduplication – The act of comparing the latest backup copy with the previous backup copy. The latest copy is kept in its entirety. In the previous copy, only the byte-level differences are kept. Although many files may have changed, at the byte level, typically only about two percent of change is encountered. Disaster Recovery – In multi-site systems, in the event of a catastrophic loss of an ExaGrid site, the ability to recover backup data that has been replicated to a surviving ExaGrid site. ExaGrid Server - The hardware on which the ExaGrid Software runs. There are several ExaGrid Server models, each with a different capacity. ExaGrid Site - A logical collection of ExaGrid Servers. An ExaGrid System can be made up of one or more ExaGrid Sites. Head – The latest back end copy of a back up. Hub – ExaGrid Systems use a hub-and-spoke topology. In multi-site ExaGrid Systems, one site is designated the hub. When logged into the hub, the hub and all spoke sites are displayed in the navigation tree and system-level reports roll up information from all spokes. When logged into a spoke, only the spoke site and its hub are displayed in the navigation tree. system and site-level reports roll up information for the spoke site only. InstantDR™ Share - A copy of a share at a second ExaGrid Site. When you create an ExaGrid Share you have the option of also creating a InstantDR Share on an ExaGrid Server in another ExaGrid Site. InstantDR Shares are used in disaster recovery operations, remote tape copy operations, and in testing data recovery from another ExaGrid Site. Landing space – The portion of an ExaGrid Server’s disk space dedicated to incoming backup data. The size of the landing space varies by ExaGrid Server model. The landing space is used to store the most recent version of your backup(s). They are stored in their entirety, undeduplicated. Since most restores are done from the most recent backup, the landing space ensures rapid restores. Replication - A share setting that instructs the ExaGrid System to maintain a second copy of a share on another ExaGrid Site. Retention Space – The portion of an ExaGrid Server that stores and manages deduplicated “Versions” of backup files.

viii

ExaGrid Systems

Share - A mount point/folder on an ExaGrid Server to which backup applications write. Share Migration - To move a share and its contents from one ExaGrid Server to another. Spoke – ExaGrid Systems use a hub-and-spoke topology. In multi-site ExaGrid Systems, one site is designated the hub. When logged into the hub, the hub and all spoke sites are displayed in the navigation tree and system-level reports roll up information from all spokes. When logged into a spoke, only the spoke site and its hub are displayed in the navigation tree. system and site-level reports roll up information for the spoke site only. TempDR Share - TempDR Shares are built manually from shares that are: replicated to a second ExaGrid Site but were created without a corresponding InstantDR Share. TempDR shares are used for test recover and in disaster recovery operations. Versions – The deduplicated copies of a backup job.


ix

x

ExaGrid Systems

The ExaGrid System Your ExaGrid System is a disk based backup product that sits behind your existing backup server. ExaGrid uses byte-level data de-duplication technology that allows your backups to be stored in a very small storage footprint. In addition, the ExaGrid System works in concert with backup applications to duplicate ExaGrid disk-based backups on to tape. This capability is especially useful to ExaGrid users who: ■ ■

Need a disaster recovery solution for an ExaGrid Single Site configuration Are required to maintain backups on removable media

This chapter provides a high-level overview of the ExaGrid System’s features and capabilities. Consult the ExaGrid online help for detailed descriptions and step-bystep instructions on how to use the ExaGrid System. Caution – Do not install third party software on your ExaGrid Server. Installing third party software may result in a loss of data and void your ExaGrid Warranty. ExaGrid Servers differ in capacity, physical size and network connectors. All ExaGrid Servers run the same ExaGrid software and may be commingled in an ExaGrid Site.

1

ExaGrid Server Configurations Model

EX1000

EX2000

EX3000

EX4000

EX5000

EX7000

EX10000E

EX13000E/SEC

Chassis

2U

2U

2U

2U

3U

3U

4U

3U

Max Single Backup

1TB

2TB

3TB

4TB

5TB

6.5TB

10TB

13TB

Disks

7

11

9

11

13

16

23

16

Disk Size

500GB

500GB

1TB

1TB

1TB

1TB

1TB

2TB

Memory

4 GiB

8 GiB

8 GiB

8 GiB

8 GiB

8 GiB

12 GiB

16 GiB

NICs

Two 1GbE

Two 1GbE

Four 1GbE

Four 1GbE and two optional 10GbE



Six 1GbE and optionally replace two 1GbE with two 10GbE

Six 1GbE and optionally replace two 1GbE with two 10GbE

All Servers

■ ■

Intel Xeon Quad-core processors ■ Enterprise SATA drives ■ RAID6 with hot spare Redundant power supplies ■ ExaGrid software ■ 19-inch rack mount chassis

■

Redundant fans

Power Requirements EX1000 through EX7000

EX10000E and Ex13000E/SEC

Idle: 230W (2.0A @115VAC)

Idle: 420W (3.6A @115VAC)

Spin up: 520W (4.5A @115VAC)

Spin up: 865W (7.5A @115VAC)

Busy: 265W (2.3A @115VAC)

Busy: 460W (4.0A @115VAC)

All ExaGrid Servers

• • • •

2

ExaGrid Systems

Frequency: 50-60 Hz Voltage: 100-240 VAC Two NEMA 5-15R 3-prong receptacles per appliance Country kits available for international orders

Environmental Requirements The following environmental requirements apply to all ExaGrid Servers: Temperature - All ExaGrid Servers

Humidity - All ExaGrid Servers

10o to 35oC (50o to 95oF)

20% to 80% Non-condensing

Cooling

• EX1000 through EX7000: 920 BTU/Hr (Busy) • EX10000E and EX13000E/SEC: 1600 BTU/Hr (Busy)

ExaGrid Hub Sites and Spoke Sites The ExaGrid System uses a hub and spoke topology where there is a single ExaGrid Site known as the hub, connected to one or more ExaGrid Sites known as spokes. Note – ExaGrid hub sites can replicate shares to spoke sites. ExaGrid spoke sites can replicate shares to their hub site. ExaGrid spoke sites cannot replicate shares between each other. When logged into an ExaGrid Server on a hub site you can: ■

■ ■ ■

See the hub site and all spoke sites displayed in the ExaGrid user interface’s navigation tree. The hub site is always listed as the top-most site in the navigation tree. View reports for the hub site and all spoke sites View events for the hub site and all spoke sites Perform hub site related management tasks

When logged into an ExaGrid Server on a spoke site you can: ■

■ ■ ■

See the spoke site and the hub site displayed in the ExaGrid user interface’s navigation tree. Other spoke sites in the system are not displayed. View reports for the spoke site View events for the spoke site Perform spoke site related management tasks

ExaGrid Administrator’s Guide 3

Data Flow The following diagram will help you follow the data flow descriptions presented in this section:

ExaGrid System Hub or Spoke Site

Hub or Spoke Site

Data flows into an ExaGrid Share via a backup application. The first time a backup job writes to a share: ■ ■

Back up application writes to an ExaGrid share. Since there are no previous backups, no byte-level data deduplication will take place.

Note – Because no byte-level data de-duplication can takes place the first time a backup job writes to an ExaGrid share, deduplication related reports will not report any significant deduplication ratios. The next time the backup job runs: ■ ■ ■

4

ExaGrid Systems

Backup application writes to the ExaGrid share again. The previous copy of the backup job is compared to current backup copy. Only the differences in the previous copy are saved. The saved set of differences from the previous backup copy is known as a “version.”

When a backup application requests a restore: ■

■

Since most restores are done from the most resent backup, the ExaGrid system keeps the most resent backup in its entirety. This ensures extremely fast restores. Restores from an older backup are done by re-assembling the backup from backup versions.


NetWorking Depending upon the model, ExaGrid Servers come with two, four or six network interface connectors (NICs). NIC

Description

NIC1

Used for communication between ExaGrid Servers. Not used

in single server, single site systems. NIC2

Used as the primary network connection for all ExaGrid Server

models. Over this connection flows: Backups, notifications, and browser sessions. NIC3, NIC4, NIC5 and NIC6

Additional NICs available on some ExaGrid Server models. Multiple NICs allow you to increase your backup through put.

Network Requirements This section details the network requirements for ExaGrid Sites and ExaGrid ExaGrid Servers. You may connect all NICs of all ExaGrid Servers in a single ExaGrid Site to the same subnet. The internal ExaGrid network (NIC1) over which ExaGrid Server communicate, can share the same subnet as the other ExaGrid NICs over which backup data flows. Network Security

The ExaGrid internal network NIC1 port is protected by explicit routes and firewall entries in the ExaGrid Servers. The ExaGrid backup NICs are protected by configurable share access lists. ExaGrid NIC1

6

ExaGrid Systems

ExaGrid Servers use their NIC1 port to communicate with other ExaGrid Servers in an ExaGrid Site and with ExaGrid Servers in other ExaGrid Sites. ■ Single-ExaGrid Server single-ExaGrid Site systems do not use their NIC1 port and can be left unconnected. ■ ExaGrid Servers within an ExaGrid Site must connect their NIC1 ports to the same subnet. ■ ExaGrid Servers within one ExaGrid Site cannot connect their NIC1 ports to the same subnet used by NIC1 ports of ExaGrid Servers in another ExaGrid Site. Failure to observe this requirement may cause loss of access to data. ■ NIC1 cannot be used to accept backups from your backup application. Multicasting Considerations for ExaGrid NIC1 ■

Multicasting is used only between ExaGrid Servers in an ExaGrid Site. Via the NIC1, multicast packets must not be routed between ExaGrid Sites. When setting up a subnet for an ExaGrid Site’s NIC1 port, ensure that standard multicasting protocols such as IGMP/IGMP snooping or multicast routing protocols like DVMRP, PIM, and MOSPF are disabled on the subnet used by your ExaGrid system. For example, Cisco 6500 users, if IGMP snooping is configured on the same subnet to which ExaGrid Servers connect their NIC1 ports, it will not allow multicast packets between ExaGrid Servers. To disable this functionality in IOS:


Select a subnet interface: Router(config)# interface subnet subnet_ID

Disable IGMP snooping: Router(config-if)# no ip igmp snooping

ExaGrid Backup NICs (NIC2 and Optional NIC3, NIC4, NIC5 and NIC6)

ExaGrid Servers use Backup NICs to accept backups from your backup servers. NIC2 must be enabled before optional NICs 3 and 4 can be enabled. The following figure details a typical ExaGrid network topology:

Site A

Site B

192.169.99.100

NIC 1

192.169.199.100 192.169.199.101

ExaGrid

NIC 2

192.169.99.101

Server 1

NIC 3

192.169.99.102

192.169.199.102

NIC 4

192.169.99.103

192.169.199.103

192.169.99.104 NIC 1

ExaGrid

NIC 2

Server 2

NIC 3

192.169.99.105 192.169.99.106 192.169.99.107

NIC 4

FIGURE 1

8

NIC 1: Each site must be on a separate subnet.

ExaGrid Systems

NICs 2,3,4,5,6: Addresses must not overlap either sites' subnet. Typically these are on the same subnet as your backup server(s).

ExaGrid Network Topology

192.169.199.104 192.169.199.105 192.169.199.106 192.169.199.107

NIC 1 NIC 2

ExaGrid

NIC 3

Server 1

NIC 4

NIC 1 NIC 2

ExaGrid

NIC 3

Server 2

NIC 4

Power Conditioning Each ExaGrid System requires conditioned power to protect it from unanticipated power events. For details on how to configure the power supplied to your ExaGrid System, consult the Assembling and Initializing Your ExaGrid System guide (PN 210-0164-xx) that came with your ExaGrid System.


ExaGrid System Overview The ExaGrid System is made up of physical and virtual storage components. Each component, whether virtual or physical, is represented in the ExaGrid System’s user interface. ■

Virtual Components ■

■

■

■

ExaGrid System — The ExaGrid System represents a network of interconnected sites. Hub Site — A hub site can be made up of one or more ExaGrid Servers. Every ExaGrid System has one hub site. Spoke Sites — A spoke site can be made up of one or more ExaGrid Servers. An ExaGrid System can have multiple spoke sites.

Physical Component ■

ExaGrid Server— Rack mounted devices for the ExaGrid System. Users and applications access the ExaGrid System through ExaGrid Servers.Single Site ExaGrid Systems

Single-Site ExaGrid Systems A single Site configuration: ■ ■

■

10 ExaGrid Systems

Is made up of one or more ExaGrid Servers. Does not have the extra protection against a full site disaster afforded by a multiSite configuration. Can be joined with another site to form a multi-site system.

Multi-site ExaGrid Systems A multi-site configuration: ■ ■ ■ ■ ■

■

Has one site designated as the hub site and one or more spoke sites. Each site can have one or more ExaGrid Servers. Allows the hub site to copy (replicate) its shares to any spoke site Allows a spoke site to replicate its shares to the hub site. Has full site disaster recovery capabilities - If one site is lost, data from a surviving site can be used to reconstruct the lost backups. Can not be split into single-site systems.


Shares Shares are mount points created on ExaGrid Servers. An ExaGrid System uses the following types of shares: 1. Backup Application Specific Shares Application specific shares are tailored to specific backup applications. ExaGrid provides detailed step-by-step instructions on how to use your backup application with these shares: a. In the ExaGrid main menu, click Help . b. From the drop down menu, select the Online Library. The Online library is displayed. c. Scroll down until you see the guide that applies to your application. d. Right mouse click on the link to the guide and save or print it as needed. To create an application specific share. See “Creating Application Specific and Application Independent Shares” on page 19.

12 ExaGrid Systems

2. Backup Application Independent Shares The ExaGrid System provides share types that, instead of being aligned with a specific backup application, are aligned with a more generic data type. Data is copied to these shares and versions of the data is managed and retained either manually or via a script. These shares include: ■

■

■

VMware Shares - Designed to accept VMDK files. VMDK files are copied to these shares either via Vizioncore vRanger Pro 3.x, manually or via a script. TAR Shares - Designed to accepted POSIX compliant TAR files copied to the share. Utility Shares - Designed to accept any type of data. Data contained in a Utility share is not compressed or deduplicated.

To create an application independent share. See “Creating Application Specific and Application Independent Shares” on page 19.


3. InstantDR™ Shares InstantDR Shares are copies of shares that are replicated to a second ExaGrid Site. InstantDR shares can be used to: ■

■

■

■

Recover replicated data after a disaster scenario. In multi-site ExaGrid Systems, backup data in ExaGrid Shares that are set to replicate to a second site, can be recovered. Test recovery of replicated data. You can use InstantDR shares to prove that the data replicated to a second ExaGrid Site is recoverable. Perform remote tape copy operations. Instant DR shares may also be used to make a tape copy of data that has been replicated to a second ExaGrid Site. For details, see the ExaGrid guide that describes how to use your backup application. A copy of your backup application guide can be found in the Online Library. Can be used with Symantec’s OpenStorage to achieve: ■ ■

■

■

■

■

Seamless integration with the ExaGrid disk backup system Unified control and tracking of all copies of data via the Symantec application’s interface -- including copies replicated to disaster recovery sites and tape copies Simplified disaster recovery, as Backup Exec is already aware of all copies of data and individually tracks retention periods for each copy Fast recovery times due to the combination of Symantec’s and ExaGrid "InstantDR" capability Short backup window, by load balancing across Symantec media servers into the ExaGrid System Flexibility to retain different amounts of backup data at a local site versus an off-site disaster recovery location, by better utilizing the unique ExaGrid architecture.

To see how InstantDR shares are created, see “Creating Application Specific and Application Independent Shares” on page 19.

14 ExaGrid Systems

4. TempDR Shares TempDR Shares are built manually from shares that are: replicated to a second ExaGrid Site but were created without a corresponding InstantDR Share. Note – In most cases InstantDR will meet your test recover and disaster recovery requirements. You will need to use TempDR shares only in those rare situations for replicated shares that do not have corresponding InstantDR shares. A TempDR Share is a “snap shot” of the replicated contents of a share that does not have a corresponding InstantDR Share. New data can not be written to them. As the name implies, they are meant to be temporary. TempDR shares are not deduplicated or replicated and occupy space on your ExaGrid System. TempDR shares are used for: ■

■

Test recovery operations: You can use TempDR shares to prove that the data replicated to a second ExaGrid Site is recoverable. Disaster recovery: To recover data while you are waiting for the lost site to be recovered.

To create a TempDR Share, see “Creating TempDR Shares” on page 22.


Share Protection Shares, and the data they contain, are protected in two ways: 1. When a share is created, you specify IP addresses of machines with write access to the share. Only backup applications and users logged into the specified machines are allowed to write to the share. 2. Multi-site configurations provide an additional level of data protection. Copies of backup files can be kept on the second, remote site. In case of a catastrophic event where one site has been damaged or destroyed, copies of backups on the second site can be used to rebuild your backups. Note – ExaGrid hub sites can replicate shares to spoke sites. ExaGrid spoke sites can replicate shares to their hub site. ExaGrid spoke sites cannot replicate shares between each other.

Share Migration Over time, your backup requirements will change. Because of these changes you may find it necessary to expand your ExaGrid System and/or add and move shares from one ExaGrid Server to another. To help you balance your backup load, the ExaGrid System allows you to move shares from one ExaGrid Server to another. This process is known as Migration. Note – Once a share has been migrated, backup jobs writing to the share have to be adjusted to accommodate the share’s new location.

16 ExaGrid Systems

The ExaGrid User Interface ExaGrid presents a unified, browser-based user interface for all ExaGrid Sites and ExaGrid Servers in an ExaGrid System. A single login gives you reports on, and administrative access to, all ExaGrid Sites and ExaGrid Servers in your ExaGrid System. The following screen capture and table detail the main components of the ExaGrid user interface:. 2

1

3

Call out

Description

1

Navigation tree: A tree view of the devices in the ExaGrid System. The hub site is always displayed as the top-most site. The navigation tree also displays a device’s status.

2

Main menu: Drop down menus that support functions for the currently selected device. Drop down menus change depending on the device selected. When logged into a spoke site, if you select the hub site or one of its ExaGrid Servers, the Manage option in the main menu is not displayed.

3

Work area: Used to display reports, configure properties, and manage the ExaGrid System.


Logging In You can access the ExaGrid user interface through any web browser on your network. To log in to the ExaGrid user interface: 1. Open a browser and enter the URL or IP address of a site’s ExaGrid Server. You can specify any ExaGrid Server but the one closest to your location will have the best performance. For example: http://ExaGridServerName or http://192.168.55.3 2. Press Return . The ExaGrid Login screen is displayed. 3. Enter your ExaGrid Administrator account’s user name and password. 4. Click OK and the ExaGrid user interface is displayed.

18 ExaGrid Systems

Creating Application Specific and Application Independent Shares Caution – Never create or delete ExaGrid shares on an ExaGrid System with external commands or utilities. Always create and delete ExaGrid shares from the ExaGrid user interface. To create a share: 1. In the navigation tree, select the ExaGrid Server on which you want to create the share. 2. In the main menu, click Manage. 3. From the drop down menu select Shares . The ExaGrid Shares page is displayed


4. Under the Tasks heading, click New. The Create New Share page is displayed .

5. From the drop down menu, select the share type 6. In the space provided, enter a name for the share. You must use unique share names on each ExaGrid Server. In addition, ExaGrid recommends using unique share names across your ExaGrid System. Doing so will: ■

Make ExaGrid reports easier to interpret

■

Avoid share name conflicts in disaster recovery scenarios

Avoid share name conflicts if you decide to migrate a share to another ExaGrid Server. 7. Select the protocol the application backup server will use to write to the ExaGrid Share: CIFS/SMB, NFS or OST. 8. If creating a Utility share, select the amount of space assigned to the share. 9. If this is a multi-site configuration, you may specify an ExaGrid Site to which this share will replicate its data: select Replicate data for this share to: Doing so puts an additional copy of the share’s contents on another site. Putting a copy onto another site provides an additional layer of protection for your data. ■

For example, you can create a share dedicated to full backups and keep a copy of those backups on the other site. At the same time you can create a share dedicated to incremental backups and not copy the share to the other site. Tip – For hub sites, consider replicating all shares to the same spoke. Doing so will make a disaster recovery easier and faster. 10. If the share is being replicated to another ExaGrid Site you may select a specific ExaGrid Server on the other ExaGrid Site on which the InstantDR™ share will be created. Doing so allows you to balance where data on the other ExaGrid Site is being placed. From the drop down list, select an ExaGrid Server on the other ExaGrid Site.

20 ExaGrid Systems

Note – If your backup application supports Open Storage Technology (OST), InstantDR Shares allow you to take full advantage of OST functionality. 11. In the spaces provided, enter the Share Access Whitelist information: a. In the spaces provided enter the IP address or DNS names of backup servers allowed to backup to this share. If a backup server uses multiple IP addresses, all IPs for that server must be entered in the list. b. In the spaces provided, enter the subnet mask associated with the backup servers that are allowed to write to this share. If you have specified an IP address for the backup server(s), leaving this filed blank will restrict access to the specified IP address. 12. To add the IP addresses and subnet masks, click Add. 13. Click Create .The new share confirmation screen is displayed. 14. Click OK and the new ExaGrid Share is created. Note – If you create a share and then some time later mark the share to replicate to a second ExaGrid Site, the “age” of the share’s replication queue is based on when the share was created and not when it was marked to be replicated.


Creating TempDR Shares To create a TempDR Share, the ExaGrid Site that hosts the source ExaGrid Share must not be reachable by the ExaGrid Site to which it replicates. The ExaGrid Site’s unavailability may be the result of a disaster scenario or in the case of a test recover operation, the result of you manually disconnecting the host ExaGrid Site’s NIC1 connection from the network. 1. Via your web browser, log into the ExaGrid Site to which the source ExaGrid Share replicates. 2. In the ExaGrid navigation tree, select the ExaGrid Server that hosts the source ExaGrid Share. 3. In the main menu click Manage 4. From the drop down menu select Shares the ExaGrid Share page is displayed 5. Select the ExaGrid source share from which you will create the TempDR Share. 6. In the task area of the page click Create TempDR Share. The Create TempDR Share page is displayed. 7. Select the ExaGrid Server on which you want to create the TempDR Share. 8. Click Create and a TempDR Share with the same name as the source share is created.

22 ExaGrid Systems

ExaGrid Reports Because the ExaGrid System employs a unified user interface, via a single log in, you can view reports for all ExaGrid Servers in your ExaGrid System. ExaGrid reports use the following icons: ExaGrid System

ExaGrid Site ExaGrid Server ExaGrid Share

Backup Job Instance. NOTE: Some backup applications create multiple instances for a single job. Tape File Replication and/or deduplication is in progress Replication and/or deduplication is complete

Backup Job

Note – The ExaGrid System is not a real time reporting system. Most reports are updated every fifteen minutes. If a report displays data that appears to be out of date, you may be viewing data between update cycles Report data is rolled up according to what is selected in the navigation tree. When you select: ■ ■

■

The ExaGrid icon, you will see reports on the entire ExaGrid System. An ExaGrid Site name, you will see reports on all ExaGrid Servers in the selected ExaGrid Site. An ExaGrid Server, you will see reports specific to the selected ExaGrid Server.

Note – All ExaGrid reports may be exported to a file in XML format.


Summary Reports A summary report is the default report displayed when you select an item in the ExaGrid navigation tree.

Summary reports provide rolled up, high-level details on: ■ ■ ■

Landing Space Retention Space Deduplication

Landing Space Summary The ExaGrid System allows you to write your backup directly to disk, at the speed of disk, without any preprocessing. This approach requires the ExaGrid System to provide a dedicated area to write backup data. The size of this dedicated area or “landing space” varies by ExaGrid Server model. The larger the ExaGrid Server, the larger the landing space. This section of the Summary report allows you to determine whether or not there is enough landing space available on an ExaGrid Server for its next backup. In most cases the space available will be 100% of the system's rating; even if a backup job was just written to the ExaGrid Server. This is because ExaGrid Servers can dynamically allocate space from their retention space on an as needed basis. As long as your backup job size does not exceed your ExaGrid System's rating, your next backup will fit.

24 ExaGrid Systems

Note – Do not confuse landing space with the total ExaGrid System space. Landing Space size is determined by ExaGrid Server model number. It is the maximum backup size for which the ExaGrid Server has been rated. Your ExaGrid Server has much more storage space than what is reported in the landing space section of this report.

Orange Landing Space Indicator If an ExaGrid Server has retained a large number of backups, there may not be enough retention space to allocate to the landing space. In this situation, if this report is viewed while the ExaGrid Server is deduplicating a previous backup job, the space available may be less than 100% and a portion of the bar graph will turn orange to indicate how much space is used by the previous backup. As processing progresses, the space available will return to 100%.

Red Landing Space Indicator If you begin to out grow an ExaGrid Server, the retention space may begin using space from the landing space. In this situation, the landing space available may be less than 100% and a portion of the bar graph will turn red to indicate how much space is now used as retention space. When landing space is used as retention space, it cannot be reclaimed until one of the following things happen: ■ ■ ■

The amount of retained data is reduced Shares are migrated to a different ExaGrid Server Additional ExaGrid Servers are added

When landing space is consumed by retained data, to determine if your next backup job will fit on the ExaGrid Server, compare the amount of landing space available with the amount of space required to accommodate the next backup.

Retention Space Summary Retention space is the amount of space dedicated to retaining versions of your backups. The amount of retention space varies by ExaGrid Server model. The amount of retention space consumed is determined by: ■ ■ ■

Size of backups The number of backups retained How long backups are retained ExaGrid Administrator’s Guide 25

■

How well each backup deduplicates

If your backup requirements change and you begin to out grow the ExaGrid Server, the retention space may begin using space from the landing space. In such cases you should evaluate the amount of data you are retaining and whether or not you need to retain less data or add capacity to your ExaGrid System.

Orange Retention Space Indicator As your ExaGrid System begins to retain deduplicated versions of backups, the space consumed by those retained backups will be represented by the “Orange” portion of the Retention Space Indicator. The orange portion of the indicator will continue to grow until your backups reach their defined retention period. At that time, the growth of the orange portion of the indicator will stabilize as new versions of your backups are retained and your backup application removes versions that exceed its defined retention period.

Red Retention Space Indicator If you begin to exceed the rating of your ExaGrid System i.e. writing 5.3TB of data to an EX5000, the ExaGrid Server will try to accommodate the backup by borrowing space from its retention space. That “borrowed” space will be represented in red on the Retention Space Indicator. In such cases you must consider either restructuring your backup jobs or adding more capacity to your ExaGrid System.

Deduplication Summary This section provides a high-level view of how well your data is deduplicating. The deduplication ratio shows how well your backups are deduplicating. The deduplication ratio is based on the total amount of backup data written and the amount of space actually consumed after deduplication. Note – The first time a backup job to writes to a share, it has no previous versions to be compared to and therefore will not be deduplicated. It will, however, be compressed. Each time a backup job runs, you will see increasingly better data deduplication ratios.

26 ExaGrid Systems

Replication Reports Replication reports provide detailed information about backup data replicating between ExaGrid Sites. Depending on the backup application, replicated data may be reported by: ■ ■ ■ ■

Server Share Backup Job Backup Job Instance

or by: ■ ■ ■ ■

Server Share Backup Job File collection - A file collection is a set of backup files that have landed on an ExaGrid Server during a specified period of time.

You may view replication reports by selecting Reports in the main menu, then from the drop down menu select a table-based report of: Jobs In Progress ■ Completed Jobs ■

Or a graph-based report of: ■ Queue Age History By Backup Job By ExaGrid Server ■ Transfer History By Backup Job By ExaGrid Server ■ Transfer rate History By Backup Job By ExaGrid Server ■ ■

■ ■

■ ■


Table-based Replication Reports The Replication Jobs In Progress and Completed Jobs reports are displayed in a table format and displays information down to the job instance level. You may set a date range for the data reported. For example:

In this Replication Completed report, the ExaGrid Server TC153SP1OB1 has a single share named TC153MSVM. The share TC153MSVM receives backup data from two backup jobs: 1. Client: ms-vm14 Agent: Windows 32-bit File System - With a single instance named: Job ID: 439 2. Client: trax6 Agent: Windows NT File System - With five job instances named: Job ID 439 through Job ID 444 From this report you can determine the amount of data transferred and when it was transferred. Large swings in the amount of data transferred may indicate that something major has changed in a backup job. A detailed description of replication tabular reports can be found in the ExaGrid online help.

Graph-based Replication Reports You can display graph-based reports for the Replication Queue History, Transfer History or Transfer Rate History for a selected ExaGrid Server(s) for the backup jobs on ExaGrid Server.

28 ExaGrid Systems

The following is an example of a Replication Queue History by backup Server report:

Graph-based reports allow you to easily see trends in your ExaGrid System’s replication. The graphs also provide controls that allow you to filter out individual items and to zoom into more detailed dates and times. A detailed description of replication graph-based reports can be found in the ExaGrid online help.


Deduplication Reports Deduplication reports provide detailed table-based information about how data in your ExaGrid System is deduplicating. Depending on the backup application, replicated data may be reported by: Server Share Backup Job Backup Job Instance

■ ■ ■ ■

or by: Server Share Backup Job File collection - A file collection is a set of backup files that have landed on an ExaGrid Server during a specified period of time.

■ ■ ■ ■

You may view deduplication reports by selecting Reports in the main menu, then from the drop down menu select either: ■ ■ ■ ■

Job In Progress Completed Jobs All Jobs by Share All jobs

30 ExaGrid Systems

For example, if you select All Jobs by Share:

In this Deduplication by Share report, the ExaGrid Server TC153SP3OB1 has a single share named TC153MSVM that receives backup data from two backup jobs, each with multiple instances: 1. Client: ms-vm14 Agent: Windows 32-bit File System 2. Client: trax6 Agent: Windows NT File System From this report you can determine the amount of data actually backed up, the amount of space it consumes on the ExaGrid Server, deduplication ratio for each job instance as well as a roll up of the deduplication ratio for all job instances. A detailed description of replication table-based reports can be found in the ExaGrid online help.


Monitoring Your ExaGrid System As detailed in “ExaGrid Reports” on page 23 the ExaGrid system allows you to manually log in and view the status of your ExaGrid System. In addition, the ExaGrid System also provides several automated tools that monitor and report status.

SMTP Email Notifications You may configure the ExaGrid System to automatically send you or others in your organization an email when the ExaGrid System generates an event. You may also configure the ExaGrid System to send a daily status report. The status report consists of a summary of the ExaGrid System’s landing space, retention space deduplication and optionally detailed backup job information. The daily status report may be sent at a specific hour of the day, to several individuals in your organization, and may be sent in HTML, or text format. Setting the format to NONE will disable sending the status report to the associated email address. To display the SMTP Email configuration page: 1. 2. 3. 4.

From From From From

the Navigation Tree, select an ExaGrid Site. the main menu, click Manage. the drop down menu, select Alert Configuration . the secondary menu, select SMTP Email. The SMTP page is displayed.

The online help has detailed instructions on how to setup SMTP Email alerting.

32 ExaGrid Systems

SNMP Traps You may also configure your ExaGrid System to sent alerts about events in the ExaGrid System via SNMP V1 traps. To display the SNMP Trap configuration page: 1. 2. 3. 4.

From the Navigation Tree, select an ExaGrid Site. From the main menu, click Manage. From the drop down menu, select Alert Configuration . From the secondary menu, select SNMP Traps. The SNMP Traps page is displayed.

The online help has detailed instructions on how to setup SNMP Trap alerting. If you plan to use an SNMP management application, such as Hewlett-Packard’s OpenView, you can download the ExaGrid Management Information Base (MIB): 1. In the ExaGrid user interface’s Main menu click Help. 2. From the drop down menu select Online Library. The ExaGrid Online Library is displayed. 3. Scroll down and right-click on ExaGrid Grid SNMP Management Information Base (MIB) and save the MIB file to your required location.

Status Monitor Your ExaGrid System has a “phone home” capability. The Status Monitor allows you to configure your ExaGrid System to send system information back to ExaGrid Support. Reports are ExaGrid Site specific. Customers with multiple ExaGrid Sites can set up status monitoring for each of their ExaGrid Sites. ExaGrid Support will automatically run tools against the report. Potential issues can be identified long before they become a problem. You may send reports to ExaGrid via FTP or email. To configure the Status Monitor for and ExaGrid Site: 1. From the Navigation Tree, select an ExaGrid Site. 2. From the main menu, click Manage.


3. From the drop down menu, select Status Monitor. The Status Monitor page is displayed.

ExaGrid MIB The ExaGrid MIB is made up of industry standard MIB and ExaGrid specific MIB objects. A copy of the ExaGrid MIB file can be found in the ExaGrid Online Library. To save a copy of the ExaGrid MIB file: 1. 2. 3. 4.

From the main menu, click Help . From the drop down menu, click Online Library. The online library is displayed. Scroll down the library entries and right click on the ExaGrid MIB file. From the pop up menus, click Save as... and save the MIB file with a name and to a location as needed

Note – Before you can use the ExaGrid MIB, you must enable SNMP on each ExaGrid Server you want to monitor. To enable SNMP on an ExaGrid Server: 1. From the ExaGrid navigation tree, select an ExaGrid Server

34 ExaGrid Systems

2. From the main menu select Identification . The Identification page is displayed:

3. On the Identification page configure the following SNMP fields: Field

Description

SNMP location

The MIB-2 system group sysLocation

SNMP System Contact

The MIB-2 system group sysContact

SNMP v2c Read Community String

Must be set to Public

Enable Access to SNMP Agent

Must be checked to enable SNMP reporting for the ExaGrid Server.

Standard MIB Objects The ExaGrid MIB supports the following “standard” MIB objects: ■ ■

■

System group from RFC 1213. See table below. The entire “ifMIB” Module (Interface MIB) – this includes 64 bit counters useful for network utilization calculations. Host resources The hrStorage table is implemented. This table includes memory statistics including total memory available (hrStorageSize) and memory used (hrStorageUsed) ■


■

The Host Resources hrProcessorLoad MIB object is implemented for all processors. This MIB object contains: The average, over the last minute, of the percentage of time that this processor was not idle.

ExaGrid Model

sysObjectId Name

sysObjectId Value

sysDescr Value

EX-1000

exagridEx1000

exagridExX000Series.1

EX1000

EX-2000

exagridEx2000


EX2000

EX-3000

exagridEx3000


EX3000

EX-4000

exagridEx4000


EX4000

EX-5000

exagridEx5000


EX5000

EX-GW

exagridExGW


EX5000GWi

EX-9000E

exagridEx9000?

exagridExX000Series.9?

EX9000?

EX-9000S

exagridEx9000?


EX9000?

EX-10000E

exagridEx10000


EX10000

EX-13000E

exagridEx13000?


EX13000?

EX-13000S

exagridEx13000?


EX13000?

ExaGrid MIB Objects ExaGrid MIB objects are made up of WholeGigaBytes and FractionalGigabytes. A 32 bit gauge does not hold values large enough to represent the bytes counts found in an ExaGrid System. To accommodate the 32 bit limitation each data item has been broken into two 32 bit counters, one for the “whole” part and one for the “fractional” part. The “whole” part contains a scaled truncated value representing GigaBytes. The fractional part contains the remainder in bytes. For example, to determine the byte count for a landing space objects, multiply the egLandingSpaceConfiguredWholeGigabytes value by 1,000,000,000 (109) and then add the egLandingSpaceConfiguredFractionalGigabytes.

ExaGrid Object Tree The ExaGrid MIB specifies the following object tree: exagridServerData (branch) ■

36 ExaGrid Systems

exagridLandingSpace (branch) ■

egLandingSpaceConfiguredWholeGigabytes (32 bit Gauge)

■

egLandingSpaceConfiguredFractionalGigabytes (32 bit Gauge)

■

■

■

■

■

■

egLandingSpaceAvailableWholeGigabytes (32 bit Gauge)

■

egLandingSpaceAvailableFractionalGigabytes (32 bit Gauge)

exagridRetentionSpace (branch) ■

egRetentionSpaceConfiguredWholeGigabytes (32 bit Gauge)

■

egRetentionSpaceConfiguredFractionalGigabytes (32 bit Gauge)

■

egRetentionSpaceAvailableWholeGigabytes (32 bit Gauge)

■

egRetentionSpaceAvailableFractionalGigabytes (32 bit Gauge)

exagridDeduplicationRatio (branch) ■

egBackupDataAvailableWholeGigabytes (32 bit Gauge)

■

egBackupDataAvailableFractionalGigabytes (32 bit Gauge)

■

egBackupDataSpaceConsumedWholeGigabytes (32 bit Gauge)

■

egBackupDataSpaceConsumedFractionalGigabytes (32 bit Gauge)

exagridPendingDeduplication (branch) ■

egPendingDeduplicationWholeGigabytes (32 bit Gauge)

■

egPendingDeduplicationFractionalGigabytes (32 bit Gauge)

■

egPendingDeduplicationAge (timeticks)

exagridPendingReplication (branch) ■

egPendingReplicationWholeGigabytes (32 bit Gauge)

■

egPendingReplicationFractionalGigabytes (32 bit Gauge)

■

egPendingReplicationAge (timeticks)

exagridServerStatus (branch) ■

egServerAlarmState (integer) 1.– no alarms 2. – 1 or more minor alarms outstanding (warning/yellow) 3. – 1 or more major alarms outstanding (error/red)

How to Calculate ■

Retention Capacity The amount of unused retention space available is found by querying the egRetentionSpaceAvailableWholeGigabytes and egRetentionSpaceAvailableFractionalGigabytes MIB objects, scaling the whole portion and then adding the fractional portion.

■

Deduplication Ratios Query all 4 objects in the exagridDeduplicationRatio group, scale the returned values to 64 bits and then convert available/consumed into a ratio. ExaGrid Administrator’s Guide 37

■

CPU utilization Average for the most recent minute CPU utilization is directly available as the hrProcessorLoad MIB object.

■

Memory utilization Point in time memory utilization is directly available as the hrStorageUsed column in the appropriate row in the hrStorage table. The row index may change with each SNMP agent restart, so the hrStorageType object must be queried to determine the correct index to use.

■

Network Utilization Several MIB objects supplied by the ifXTable in the ifMIB can assist in determining Network Utilization. Of particular interest are: ifHCInOctets and ifHCOutOctets. ifHCInOctets is defined as: The total number of octets received on the interface, including framing characters. This object is a 64-bit version of ifInOctets”

Note – Cisco has published an article titled How to calculate bandwidth utilization using SNMP. It is available at: http://www.cisco.com/en/US/tech/tk648/tk362/technologies_tech_note09186a008009496e.shtml

This article describes how these counters can be used over time to determine network utilization. With full duplex media, two formulas are required, one for input and one for output. The ifInOctets and ifOutOctets high capacity counters should be polled keeping track of the time between polls. The Formulas become: input utilization = change in ifInOctets * 8 * 100 / pollSeconds * ifSpeed output utilization = change in ifOutOctets * 8 * 100 / pollSeconds * ifSpeed

38 ExaGrid Systems

Restoring Files All restores are done via your backup application. From you backup application’s perspective, there is no difference between restoring from an ExaGrid System and a tape system. Consult your backup application’s documentation for details on how to restore data.


ExaGrid System Maintenance Management and maintenance tasks for your ExaGrid System include: 1. 2. 3. 4.

40 ExaGrid Systems

Shutting down and restarting ExaGrid Servers Managing ExaGrid accounts Adding additional ExaGrid Servers Record keeping

Shutting Down and Restarting an ExaGrid Server Caution – NEVER shut down or restart an ExaGrid Server by cutting power or pressing the power or restart buttons on the front panel. ExaGrid Servers must be shutdown or restarted via the ExaGrid user interface. To shut down an ExaGrid Server: 1. In the navigation tree, select the ExaGrid Server to be shut down. 2. From the main menu, click Manage . 3. From the drop down menu, click Administrative State . The Administrative State page is displayed. 4. Click Shutdown . To restart an ExaGrid Server: 1. In the navigation tree, select the ExaGrid Server to be restarted. 2. From the main menu, click Manage . 3. From the drop down menu, click Administrative State . The Administrative State page is displayed. 4. Click Restart.

Account Management Admin accounts are used to access the ExaGrid System. The ExaGrid Manager Admin account comes with a default password. The account name and password were given to you when your system was installed. ExaGrid recommends that you change the default passwords.


Admin Accounts Use the ExaGrid Manager Admin account to perform management tasks and view reports. The default user name is admin. The default password was given to you when your ExaGrid System was installed. When logged into an ExaGrid Manager Admin account, you can manage all devices in the Site.

Changing ExaGrid Admin Passwords To change an ExaGrid Admin account password: 1. From a web browser, enter the URL or IP address of a site’s ExaGrid Server. For example: http://ExaGridName or http://192.168.55.3. You can specify any ExaGrid Server but the one closest to your location will have the best network performance. 2. At the log in prompt, enter the Admin user name and password and click OK. 3. From the navigation tree, select an ExaGrid site. 4. In the main menu, click Manage . 5. From the drop down menu, select Administrators . 6. In the right-hand frame, select the Admin account and click Modify Login. 7. Enter a new password in the Password box. Passwords can contain the following alphanumeric characters: A-Z or a-z, 0-9, Underscores ( _ ) 8. Click Apply. The new password will be implemented.

42 ExaGrid Systems

Adding Additional ExaGrid Servers As your backup requirements grow, you may find it necessary to add additional ExaGrid Servers to your ExaGrid System. Note – Step-by-step instructions on how to install and initialize a new ExaGrid Server can be found in the Assembling and Initializing Your ExaGrid System guide that came with your new ExaGrid Server. When you add an ExaGrid Server, you not only increase the amount of storage space in your ExaGrid System, you also increase the system’s ability to take in and process backup data. More ExaGrid Servers mean more memory, CPUs and ports over which backup data can flow. When a new ExaGrid Server is added, the ExaGrid System will automatically redistribute retained backup versions to balance the storage load across the ExaGrid Servers. Note – Adding a new server will not automatically move (migrate) shares to a new server. Share migration is a manual process. To take advantage of the additional ports on the new ExaGrid Server(s): 1. Evaluate existing shares and the jobs writing to them to identify which shares, if moved (migrated) to the new ExaGrid Server, would improve your backup through put. 2. Migrate the share(s) to the new ExaGrid Server(s). 3. Redirect backup job(s) to the new share location. Note – Consult the ExaGrid System’s online help for instructions on how to migrate a share.


Adding an ExaGrid Site to an ExaGrid System The ExaGrid System uses a hub and spoke topology. Every ExaGrid System has one hub site. A single-site ExaGrid System is a hub site. Multi-site systems have one hub and may have one or more spoke sites. In most cases, ExaGrid Hub and Spoke Sites will be designated as such during the initialization process. To add an ExaGrid Site designated as ExaGrid Spoke Site to an ExaGrid System: 1. In the left-hand left-hand navigation navigation tree, tree, select the ExaGrid ExaGrid icon. icon. 2. In the main menu, click click Manage, 3. From the drop drop down menu, select Add Spoke . The Adding a New Spoke page is displayed. 4. In the space provided, enter the IP address of the spoke to be added. To find the IP address of the ExaGrid Site, via a web browser, navigate to an ExaGrid Server in the ExaGrid Site. In the ExaGrid navigation tree, place your mouse cursor over the ExaGrid Site’s name. The displayed tool tip contains the site’s IP address. For example:

In some cases you may want an existing, single-site ExaGrid System to be joined with another ExaGrid Site. To do so, via the ExaGrid user interface you must: 1. Reconfigure Reconfigure the existing single-site single-site ExaGrid ExaGrid System as an ExaGrid ExaGrid Spoke Site. Site. 2. Join the newly designated ExaGrid Spoke Site to an ExaGrid Hub Site.

44 ExaG ExaGri rid d Syst System ems s

Reconfiguring a Single-site as a Spoke Site Note – Only single-site ExaGrid Systems can be reconfigured as an ExaGrid Spoke Site. The ExaGrid Site in a single-site ExaGrid System is an ExaGrid Hub Site. To reconfigure the ExaGrid single-site as an ExaGrid Spoke Site: 1. Log into an ExaGrid ExaGrid Server on on the ExaGrid System System to be redesignated redesignated as an ExaGrid Spoke Site. 2. In the ExaGrid user interface’s navigation tree, click ExaGrid . 3. In the main menu click Manage . 4. From the drop down menu click Network . 5. From the secondary menu, click Topology. Topology. The Change Topology page is displayed. 6. Click Make Spoke . The site will be redesignated as a spoke.

Joining an ExaGrid Spoke Site to an ExaGrid Hub Site To join an ExaGrid Spoke Site to an ExaGrid Hub Site: 1. 2. 3. 4. 5. 6.

Log into into an ExaGrid ExaGrid Server Server on the the hub site. site. From the navigation tree, select ExaGrid. In the main menu, click Manage. From the drop down menu, select Add Spoke. In the space provided, enter the IP address of the ExaGrid Spoke Site. Click Add and the ExaGrid Spoke Site will be joined with the ExaGrid Hub Site.

ExaGrid Administrator’ Administrator’s s Guide 45

Record Keeping ExaGrid recommends that you keep detailed records for each share. Whether online or paper, keep the records in a secure place. The contents of a share record will aid current and future members of your organization. The following is an example of a share record.

Share Record Form Share name:

Date created:

Created by:

Date modified:

Modified by

Date deleted:

Deleted by:

Replicated to a second Site

Yes____

No____

Yes____

No No____

Backup Server Type

Type of data being backed up

Copy to tape

46 ExaG ExaGri rid d Syst System ems s

ExaGrid System Disaster Recovery In the event of a catastrophic event where an entire ExaGrid site is no longer available, the data from the lost site can be recovered if: ■ ■ ■

The lost site was part of a multi-site system The second site is still available Shares on the lost site were configured to replicate to the second site

To perform a disaster recovery, in your surviving ExaGrid System’s online library, consult: ■

■

The ExaGrid Disaster Recovery Guide. This guide provides step-by-step instructions on how to recover your lost ExaGrid System. Once the ExaGrid System has been recovered, consult the ExaGrid application guide for your backup application. This guide provides step-by-step instructions on how to recover your data via your backup application.


Troubleshooting In the unlikely event that your ExaGrid System experiences a failure, use this section to help troubleshoot that failure.

Disk Failure Depending upon the model, an ExaGrid Server chassis contains 12, 14 16 or 24 disk drives. Each ExaGrid Server has RAID 6 with a hot spare. This means that an ExaGrid Server can lose up to three disks and continue functioning. However, to ensure optimum performance, failed disks should be replaced as soon as possible. Replace the disk in accordance with the maintenance plan that came with your ExaGrid Gateway. If an ExaGrid Server disk fails, a “90524” alert will be generated indicating that the failure has occurred and which disk has failed. The alert will be sent to all users who normally receive email alerts from your ExaGrid System. The alert message will also appear in the Events report in the ExaGrid System’s user interface. ExaGrid Server disk failure alerts use the following format: Drive n “[252:3]” (Model: WD…, FW Rev.: 03.00C05 , SN: WD-…) was removed or failed. From MegaRAID84016E SN: P097070408.

Where “n” is the Physical Disk number. See “Disk Numbering” on page 49 for disk locations.

48 ExaGrid Systems

Disk Numbering The following illustrations detail disk numbers, spares and empty disk bays for each ExaGrid model. Note – The following drawings identify the location of spare disk drives when an ExaGrid Server is first installed. If a disk drive fails, the spare will take the failed drive’s place. When you replace the failed drive, the replacement drive will become the spare.

3

6

Empty

Empty

2

5

Empty

Empty

1

4

Spare

Empty

EX1000 Disk Numbering

3

6

9

Empty

2

5

8

Spare

1

4

7

10


Spare

Empty

3

6

2

5

8

Empty

1

4

7

Empty



3

6

9

Empty

2

5

8

Spare

1

4

7

10


4

8

12

Empty

3

7

11

Empty

2

6

10

Empty

1

5

9

Spare


3

6

9

Spare

2

5

8

11

1

4

7

10

EX9000E Disk Numbering

50 ExaGrid Systems

6

12

18

Empty

5

11

17

Spare

4

10

16

22

3

9

15

21

2

8

14

20

1

7

13

19


4

8

12

Spare

3

7

11

15

2

6

10

14

1

5

9

13


Power Supply Failure Depending upon the model, an ExaGrid Server chassis uses two or three power supplies. In the event of a power supply failure, the ExaGrid Server can continue functioning. However, to maintain the safety margin found in having redundant power supplies, the failed power supply should be replaced as soon as possible. Replace the power supply in accordance with the maintenance plan that came with your ExaGrid Server.


If an ExaGrid Server power supply fails, a “90530” alert will be generated indicating that the failure has occurred. The alert will be sent to all users who normally receive email alerts from your ExaGrid Server. The alert message will also appear in the Events report in the ExaGrid System’s user interface. In addition, the failed power supply’s LED status indicator will change from green to amber. ExaGrid Server power supply failure alerts use the following format: Event ID: 90530 Severity/Type: Warning Grid Name: Grid Name Site Name: Your_Site SiteGrid Name: name Device Name: Server Name Description: Power Supply Fault Detail: Power supplies are indicating a fault condition

Network Debugging Tools ExaGrid Systems are appliances. Access to an ExaGrid System is through the ExaGrid System’s web-based user interface. In the unlikely event you need to debug your ExaGrid network settings, the ExaGrid System’s user interface provides the following tools: 1. 2. 3. 4.

Ping Traceroute Network interface statistics Network performance test

To use these tools: 1. 2. 3. 4.

52 ExaGrid Systems

In the navigation tree select ExaGrid, an ExaGrid site or an ExaGrid Server. In the main menu, click Manage . From the drop down menu select Network . From the secondary drop down menu, select Diagnostics . The network Diagnostics page is displayed.

Notifications If you encounter issues with receiving email alerts or SNMP traps from your ExaGrid System: ■

■

■

If your email server is relay enabled, a missing Send from password will cause the system to send all messages as relays. Relayed messages are often times treated as spam by spam filters and filtered out. If you must configure messages as relays, ensure that the recipient’s spam filters allow relayed messages from this site. See “Network Debugging Tools” on page 52. to determine proper network connectivity for each ExaGrid Server. Double check the DNS definition for each ExaGrid Server:

1. 2. 3. 4.

In the navigation tree, select an ExaGrid Server. In the Main menu, click Manage. From the drop down menu select Network . From the drop down menu select Configuration . The Network Configuration page for the selected server is displayed. 5. Ensure the DNS IP address is correct. If it is not, enter the correct IP address and click OK. ■ For SNMP traps, ensure that you are using the latest ExaGrid MIB: 1. In the ExaGrid user interface’s Main menu click Help. 2. From the drop down menu select Online Library. The ExaGrid Online Library is displayed. 3. Scroll down and right-click on ExaGrid Grid SNMP Management Information Base (MIB). 4. Save the MIB file to your required location.

Consumption In the unlikely event that your ExaGrid System begins experiencing rapid consumption of storage space: Ensure the size of your backups and the length of time they are retained does not exceed your system’s rating. When you purchased your ExaGrid System, your ExaGrid System Engineer sized your system according to your current and future backup requirements. If those requirements have changed, it is possible that you are now be exceeding your system’s rated capabilities. Review your current backup sizes and retention periods.


Ensure your backup deduplication ratios are getting consistently better over time. Review the Deduplication report for All Backup Jobs . Each time a job runs, its deduplication ratios should improve.

The site-level Deduplication Summary report will give you a general overview of deduplication in your ExaGrid System. Use the Deduplication report for All Backup Jobs to gather more detailed information about your ExaGrid System’s deduplication. A sudden, negative change in deduplication ratios can be the result of: ■

■

■

Adding a very large percentage of new data to the backup. This includes renaming previously backed files and directories. The ExaGrid System sees renamed backup client files, directories, and all of the files contained in renamed directories, as new and they will not be deduplicated against the files and directories with the old names. Renaming backup jobs. Renamed jobs are seen by the ExaGrid System as new backup jobs and will not be deduplicated against the backup jobs with the old names. In isolated cases, upgrading the version of your backup application can cause a one time drop in your deduplication ratio. Subsequent backups with the new backup application version will give you increasingly better deduplication ratios. More and more disk space will be freed up as the retained backups written by the old backup application version are aged out.

Replication In the unlikely event that your ExaGrid System experiences issues replicating deduplicated data to another site use the following suggestions to troubleshoot the issue: ■

■

1. 2. 3. 4.

54 ExaGrid Systems

See the ExaGrid “Network Debugging Tools” on page 52 to ensure that the necessary network bandwidth is available. Review the ExaGrid Site’s replication schedule and ensure it provides a wide enough time window to replicate the ExaGrid site’s data. To view an ExaGrid site’s replication schedule: In the navigation tree, select an ExaGrid Site’s name In the main menu, click Manage . From the drop down menu, select Network . From the drop down menu, select Replication Schedule . The ExaGrid Site’s replication schedule is displayed.

■

■

■

Use the ExaGrid Server-level Replication Transfer Rate History report to identify daily replication trends. Look for changes in the replication rate. A change may indicate other network traffic is competing with the transfer of your deduplicated data. Use the Replication Transfer History by Backup Job report to identify when a backup job’s data was available to be replicated. Use the Replication History report to identify how much data each job contributes to overall replication.


Performance Tuning Your ExaGrid Server will accept data as fast as your backup server or media server can write it. Most backup application vendors provide performance tuning documentation. If you experience slower than expected writes to your ExaGrid System, ExaGrid recommends that you consult the backup vendor’s documentation to get the most out of your back up application. In general, ExaGrid has found that the following types of things, if addressed, will maximize backing up to an ExaGrid System: ■

■

■

■

■

56 ExaGrid Systems

If your backup server or media server is equipped with a network card that has off loading capabilities, off load as much processing as possible to the network card. This will free up the server’s CPU to process and write the back up faster. Review and adjust your backup server’s or media server’s network buffer size. Optimal settings will vary according to backup application and server type. NetBackup users should review their Shared memory settings (number and size of data buffers). Consult the NetBackup’s Tuning guide that came with your NetBackup backup application. NetBackup users should review their client communications buffer size. When NetBackup clients are installed, buffers are set from 16K to 64K. ExaGrid has found that 128K or larger can enhance performance. NetBackup users should review their server and client communications buffer size and ensure that they are both set to the same value.

Ex a Grid Admin Guide

Recommend Documents