Dell DR4000 Release Notes

NOTE: A NOTE indicates important information that helps you make better use of your computer. CAUTION: A CAUTION indicates potential damage to hardware or loss of data if instructions are not followed. WARNING: A WARNING indicates a potential for property damage, personal ____________________ Information in this publication is subject to change without notice. © 2012 Dell Inc. All rights reserved. Reproduction of these materials in any manner whatsoever without the written permission of Dell Inc. is strictly forbidden. Trademarks used in this text: Dell™, the DELL logo, and DR4000™ are trademarks of Dell Inc. Microsoft®, Windows®, and Windows Server® are either trademarks or registered trademarks of Microsoft Corporation in the United States and/or other countries. Symantec™ and Backup Exec™ are trademarks or registered trademarks of Symantec Corporation or its affiliates in the U.S. and other countries. Other trademarks and trade names may be used in this publication to refer to either the entities claiming the marks and names or their products. Dell Inc. disclaims any proprietary interest in trademarks and trade names other than its own. 2012 - 01 Rev. A00

1 Overview: DR4000 System Release Notes The Release Notes for 1.0 are intended to supplement the Dell DR4000 system reference documentation available at support.dell.com/manuals. The release notes contain important details about the system software, which is pre-installed on Dell DR4000 systems, and documents the related known issues, resolutions, and other important information needed for the 1.0 release of the DR4000 system. Supported Software and Hardware For a complete list of the supported hardware and software for the Dell DR4000 system, see the Dell DR4000 System Interoperability Guide available at support.dell.com/manuals. Source Code Availability A portion of the DR4000 system software may contain or consist of open source software, which you can use under the terms and conditions of the specific license under which the open source software is distributed. Under certain open source software licenses, you are also entitled to obtain the corresponding source files. You may find the corresponding source files for respective programs at http://opensource.dell.com. Checking System Information Before troubleshooting issues with any DR4000 system, ensure that the installed DR4000 system software is up to date by checking its software version against the available versions at support.dell.com/support/downloads. For information about checking the current version of the DR4000 system software installed, see "Verifying the Current Software Version" in Chapter 7 of the Dell DR4000 System Administrator Guide available at support.dell.com/manuals.

Known Issues and Resolutions This section explains the known issues, descriptions, status, and any workaround for the DR4000 system software. Table 1. Known Best Practices Issues and Workarounds for Release 1.0 Issue Description Workaround/Solution 4051 Recover, restore, and source-target replication issues. Data loss can occur when a DR4000 administrator does not disable the replication relationship between the source and target containers before using the DR4000 system command-line interface (CLI) system --init command on the source container. CAUTION: The CLI system --init command reconfigures the DR4000 system and removes all existing data, and recreates the file system to factory settings. Dell does not recommend using this command on systems containing production data because it destroys any existing data. If you must use the CLI system -- init command, make sure that you first disable the replication relationship between the source and target DR4000 systems. CAUTION: Failure to observe this practice may corrupt the data that is stored on the replication target system. 6381 Reduced savings issues encountered when compression- enabled data is ingested from DMA data sources. The DR4000 system experiences reduced savings rates (less than optimal) when the data it ingests is already compression-enabled by the native data source. It is recommended that you disable the data compression used by the data source for any first-time backups. For optimal savings, native data sources need to send data in its raw state for ingestion.

3 8878 "Manual Intervention Required" message displayed. The DR4000 system encountered an issue during the calibration of the super capacitor that caused the "Manual Intervention Required" system state message to display. After this message displays, allow the DR4000 system 15-30 minutes before attempting a cold reboot of the system. The DR4000 system should start up with all services running normally. 8897 DR4000 and iDRAC reboot mismatch. Currently, the DR4000 system can encounter some low-level system reboot management coordination issues when the reboot is initiated using iDRAC. Initiate the reboot of the DR4000 system using either of the two following choices: • DR4000 system CLI command (system --reboot) • DR4000 system GUI command (System Configuration Reboot) When the reboot is initiated using system CLI or GUI commands, the data contents are properly and fully flushed to disk, before the system is shut down. Failure to follow this practice will cause the DR4000 system to start filesystem recovery during the boot process (and lengthen the time required for system reboot to complete). Table 1. Known Best Practices Issues and Workarounds for Release 1.0 (continued) Issue Description Workaround/Solution

8909 reinit_nvram fails with: NVRAM ’write ready’ time out exceeded. NVRAM reports errors and fails memory tests when the proper procedure for powering off the DR4000 system and waiting for the NVRAM to properly discharge is not followed. Before attempting to remove or replace the NVRAM in the DR4000 system, you must observe the steps documented in "Shutting Down/Starting Up a DR4000 System" in the Dell DR4000 System Administrator Guide. Before attempting to remove or replace the NVRAM in the DR4000 system, observe the following best practice for shutting down the DR4000 system: 1 Power off the system using either the DR4000 system GUI (select Shutdown in System Configuration page, or CLI (use the system --shutdown command). 2 Allow the system to fully complete its power-down process (when the power supply status indicator is unlit). 3 Disconnect the system power cables from the electrical power outlet. 4 Ensure that you wait an additional period of time (up to 10 minutes) and/or verify that all of the green and amber NVRAM LEDs on the rear panel of the system chassis are unlit. NOTE: If you do not allow the NVRAM super capacitor sufficient time to discharge, the NVRAM status will report DATA LOSS when the system is subsequently powered on. Table 1. Known Best Practices Issues and Workarounds for Release 1.0 (continued) Issue Description Workaround/Solution

5 Known Configuration Issues Table 2. Known Configuration Issues and Workarounds for Release 1.x Issue Description Workaround/Solution 425 Reverse replication not supported on DR4000. Replica containers are always in a R-O (read-only) mode on a DR4000 system and write operations are not supported. Under certain conditions, it might be beneficial for replica containers to support write operations for restoring data from an archival target. For example, data could be replicated back to the remote site where the DMA is connected to allow this data to be restored directly. This situation applies to configurations where data is backed up from a remote location to a local container, and then replicated over a WAN to a replica container that is backed up to tape. To leverage deduplication across the WAN for this scenario: 1 Make sure that the replication operation has completed (between source and target). 2 Delete current replication relationship, and re-create replication relationship (reversing the source and target roles). 3 Restore data to the original target container (now the source). 4 Make sure that the replication operation has completed. 5 Delete replication relationship and re-create replication relationship (restoring original source and target destinations). Under this scenario, a fraction of the data to be recovered is sent across the WAN link. This could speed up a remote restore significantly.

425 Reverse replication not supported on DR4000. (cont.) The data needs to be restored from the tape backup to the original location (first back to the DR4000 replica container, and then back to the original source location of the data on the other side of the WAN link). NOTE: If you choose to use the alternate workaround method, you must set up a new data storage unit in the DMA and import the images before a restore to the original location can occur. However, there are some downsides: • If step 1 is not followed correctly, any changes not fully replicated are lost. • During steps 2 and 3, any data written to the original source container may be lost. • During step 4, if the data is not fully replicated back before the switch is made, it may be lost. Alternatively: 1 Create new container on target DR4000. 2 Set up replication from this container back to source DR4000 container. 3 Set up a new disk storage unit in DMA and make sure that the DMA is aware of new images. 4 Import old images back into the DMA from target DR4000 (the original source location). 5 Use new disk storage unit in DMA and restore data back to the original client. Table 2. Known Configuration Issues and Workarounds for Release 1.x (continued) Issue Description Workaround/Solution

7 4256 NFS: DR4000 returns an incorrect error message to the client for NFSv4 - mount/UDP (1.1). The DR4000 system only supports NFS Version 3 and TCP. Some other operating systems (such as Solaris and HP-UX) mount their containers using the UDP protocol and NFS using the TCP protocol. However, these clients are not supported by the DR4000 system. There is no workaround or solution (OS clients must support NFS Version 3 and TCP). 5601 Authenticati on failure when entering in username and password correctly. The DR4000 system reports authentication failure back to the client if it cannot contact the domain controller. If the administrator is sure that the username and password are correctly entered, log into the DR4000 CLI and enter the following command: network --ping --destination 6957 system --init failed with "Failed to unmount /dev/sbd3". This condition can occur when an administrator has enabled replications, there were active snapshots, and the DR4000 CLI commands were entered to reset the system to its factory settings: maintenance --filesystem -- wipe_data system --init During the filesystem initialization, the mkfs.vxfs are not properly unmounting all the snapshots. Use the Factory Reset option in the Restore Manager (RM) program, to reset the DR4000 system to its factory configuration. WARNING: The Factory Reset mode deletes all DR4000 system data. Factory Reset mode should only be used when the data is no longer needed. NOTE: Contact Dell Support for assistance prior to using the RM program. Table 2. Known Configuration Issues and Workarounds for Release 1.x (continued) Issue Description Workaround/Solution

7567 CommVault reports failures if restoring using CIFS when using options to ignore ACLs even when the restore operation succeeds. The DR4000 system does not support a CIFS-based access control list (ACL) native to the Microsoft Windows filesystem. If an administrator tries to restore a backup image on to a different system container, the system does not support preserving the ACL for the individual files that are part of the CommVault (CV) backup image. Some administrators attempt to avoid finding new disk space to restore an image by assuming they will get 100% deduplication and use less physical space on the disk. Administrators should not restore files on the DR4000 system that depend upon CIFS and an ACL for individual files that need to be preserved. 8028 Network restart should say whether DNS is updated or not every time (or if something changes). If DHCP is configured, the domain and IP addressing information is updated automatically. Make sure that the information received from the local DNS server is correct. This issue is magnified when replication has been configured with a hostname, but has a different domain name (for example, acme.com is the DNS suffix in DHCP and the hostname DNS suffix is acmewest.com). Check to verify that the information is correct by using the following DR4000 system CLI command: system --show 8099 Relay server does not work as IP or FQDN (use only hostname). When an administrator configures e-mail to receive system alerts from the DR4000 system, the relay host has to be configured. When configuring the relay host only the hostname should be given (since the DNS suffix is automatically appended, this is what causes the issue). Use only the hostname, and send test e-mail messages to verify that e-mails are received by all recipients. Table 2. Known Configuration Issues and Workarounds for Release 1.x (continued) Issue Description Workaround/Solution

9 8402 "replication -- update" fails if the peer container whose IP is changed is offline. This issue can occur when an administrator has configured replication to use the IP address instead of a fully qualified domain name (FQDN) for a remote DR4000 system, and the IP address has been changed because of a network topology change. The administrator needs to update the configuration database with the new IP address. This update can only take place when the remote DR4000 system is also active because this is a synchronous operation. 8449 Hostname on the appliance is not always the name configured at initialization. When an administrator configures a DR4000 system with a hostname "A", and DHCP assigns it different hostname, it is not apparent that this change has taken place. Administrators need to ensure that the configured hostname and the DHCP-assigned hostname are the same. Use the DR4000 system GUI to check the System Name in the System Information pane on the Dashboard page, or use the DR4000 CLI command: system --show Table 2. Known Configuration Issues and Workarounds for Release 1.x (continued) Issue Description Workaround/Solution

8559 system --stats do not get updated immediately for unlinks/ truncates, or overwrites. For unlinks/truncates, when a file is unlinked, the Current Bytes, Post Deduplication Bytes, and Deduplication Savings do not get updated immediately. The cleaner operation needs to complete its cycle before these values get updated. For overwrites, when a file gets overwritten, the Current Bytes reflects this change correctly except when the overwritten file is the same exact file size. However, the overwritten content is cleaned up by the cleaner operation in the background similar to how unlinks/truncates are updated with Post Deduplication and Deduplication Savings updated after the cleaner operation is completed. There is no current workaround or solution. Cleaner operations are designed to be a background process. 8583 Replication down event was not sent from source container when target entered Maintenance mode. When a target DR4000 system in a source-target pair enters Maintenance mode, no notification is sent to the administrator from the source DR4000 system about this mode change. The only workaround is to rely on system alerts and notifications from the target DR4000 system. Table 2. Known Configuration Issues and Workarounds for Release 1.x (continued) Issue Description Workaround/Solution

11 Table 3. Known Configuration Issues and Workarounds for Release 1.0 Issue Description Workaround/Solution 1994 Linux CIFS client may corrupt files due to issues with some client types. The DR4000 system does not support Linux clients mounting on CIFS. There is no workaround or solution. The DR4000 system only supports Linux clients mounting from NFS. 5957 Backup Exec restore fails with "Backup media errors." When the DR4000 system enters Maintenance mode, from the Backup Exec 2010 perspective some contents of the physical media library disappear from the backup folder. This can cause the restore operation to fail, and then cause a Backup media error message to display. After the DR4000 system completes its Maintenance mode and returns to its operational state, Backup Exec should perform the following: • Rescan the backup folder to reconnect to the media. • Restart the restore operation. 5995 A potential data issue can occur when using multiple DR4000 containers for a single CV backup, and replication is not enabled on all containers. The best practice when using multiple DR4000 containers for performing a single CommVault (CV) backup and using the DR4000 replication feature, is to enable replication for all containers in the CV backup. Make sure that replication is enabled for all of the DR4000 containers that are involved in the CV backup.

7351 File or directory creation fails with an ENOSPC error. ENOSPC means there is insufficient space on the system because: • The maximum usable capacity on the disk is reached. • The maximum supported inode limit in the system is reached. NOTE: An inode is an index-node data structure in a file system. The maximum number of inodes (files, directories, or symbolic links) supported in the DR4000 system is 4 million. File or directory creation will start failing when the number of inodes in the DR4000 system exceeds 4 million. Additional space on the DR4000 system must be created for new inodes by deleting older or unused files and directories in the system. The DR4000 system is not intended to be used as a basic file share. The system is designed to be used with DMA backup software that puts individual backup files into larger backup data chunks, which reduces the total number of inodes needed. 7790 Sending e-mail alerts to multiple recipients does not work. Using multiple recipients for e- mail alerts generates errors with the e-mail subsystem. Use a single recipient or a single distribution list. 8392 Administrator cannot allow open access to a container via NFS once it has been created. The DR4000 system reports an error because "*" is not a valid argument for --clients. The administrator needs to either delete and recreate the container if there is no data in the container, or add the lists of hosts that can access the share. This issue will be fixed in the next scheduled release of the DR4000 system. Table 3. Known Configuration Issues and Workarounds for Release 1.0 (continued) Issue Description Workaround/Solution

13 Related DR4000 System Documentation For a complete list of the released documentation that supports the Dell DR4000 system, see the latest available versions at support.dell.com/manuals. This product is supported by the following Dell documentation: • Dell DR4000 System Getting Started With Your System • Dell DR4000 System Owner’s Manual • Setting Up Your Dell DR4000 System • Dell DR4000 System Administrator Guide • Dell DR4000 System Command Line Reference Guide • Dell DR4000 System Interoperability Guide 8122 Upgrade reports 100% complete, but system may be unresponsive for 15 minutes. During a DR4000 system software upgrade, the progress indicator may report 100% completion prior to actual completion of the software upgrade. At that time, the system is applying updated settings and it may require additional time to fully complete the upgrade. The software upgrade process may continue to run for an additional 15 to 30 minutes. No user action is required. CAUTION: Do not interrupt or reboot the system during the software upgrade process. Table 3. Known Configuration Issues and Workarounds for Release 1.0 (continued) Issue Description Workaround/Solution