DR Series Software 4.0.3 - Release Notes

Upgrade to 4.0.0.1 might fail when multiple roles are assigned to the backup_user.

31848

The system cleaner is not triggered when the low space threshold is met on a 12TB DR2000V.

31844

"Post encryption bytes" is displayed incorrectly.

31817

10G cards are not recognized in the 4.0.0.1 release of the DR Series software.

31741

Various issues with editing NFS/CIFS connection offline on a container in the DR Series system GUI.

31583

Earlier versions of the software included the ability to configure arcfour256 and arcfour128.

30986

As of the 4.0.0.3 release, Arcfour ciphers are no longer accepted for ssh connections.

30440,

24597

CIFS traffic is showing up in the usage graphs when only running iSCSI.

Cannot shut down the system from the GUI.

27931

When the dictionary is being upgraded from DR Series system software version 4.0 to a later hot-fix release, it can take several minutes to a few hours (depending on the dictionary size), but no indication of status is given. Containers may not appear on the Containers page in the DR Series system GUI during the upgrade. The system should notify the user of status during the dictionary upgrade. (Note: This upgrade is a one-time action taken for upgrade to version 4.0.0.1.)

31461

This issue was addressed in the 4.0.0.1 release. Proper communication is now given with notifications of upgrade, initialization, and installation of container. Once the upgrade is complete, all containers and associated statistics are listed on the Containers page of the GUI.

31252

31126

Savings/performance issue occurs in release 4.0 of the DR Series system software after upgrade. After upgrading to 4.0, you may see a slight decrease in savings percentage on the next update; over time, this savings will come back in line as the old data stream is aged out.

Cleaner schedules that span multiple days are not retained upon upgrade to 4.0.

31114

30895

The DR Series 4.0 OST plugin does not support BE16 or NetBackup 8.0.

30964

Network graph in the GUI used "read/write" for "transmit/receive."

30910

30820

Critical alerts incorrectly generated and displayed before the DR Series system finishes initializing.

30902

Error on DR4000 after upgrade when backing up to 1.0 data store.

31243

Resolved in the 4.0.0278.0a release. This issue only impacted a DR4000 that still contained release 1.0 backup data.

31059

Loop in self-healing routine in specific failure scenario.

Resolved in the 4.0.0278.0a release.

31017

Replication error upon bringing system into maintenance mode when using older data format from a 2.x release.

Resolved in the 4.0.0278.0a release.

30375

New Quest DLV license support.

This feature was added in version 4.0 to support Quest and Dell license formats.

30440,

24597

CIFS traffic is showing up in the usage graphs when only running iSCSI.

This issue was fixed in version 4.0.

30422

CIFS Administrator password migration from DR Series system software version 3.x to 4.x.

During the upgrade to version 4.0:

Users corresponding to local workgroup users used for CIFS authentication will be created with the CIFS role and a default password (St0r@ge!).

A Message will inform the users of the above and request him/her to set the password of these users to the password of the corresponding local workgroup accounts.

For local workgroup accounts for which the migration has failed, a message will be displayed asking users to delete these accounts with "authenticate --delete" commands, and then create users with the CIFS role and appropriate CIFS authentication information manually.

29936

Domain users that are not in login group can still log on.

BE15 is unable to detect NDMP VTL devices mapped to NetApp C Mode filer when NetApp filer is configured in vserver scope mode.

29769

Configure NetApp for 7 Mode or for NetApp C Mode for node scope mode. The Node IP used is the NDMP host to the BE server.

29704

File system crash.

Creating any new username containing "administrator" substring as leading characters in the username and CIFS role causes administrator login via the GUI to be locked out.

29696

DR4300e maxing out at 4 million files on some configurations.

29672

This issue was addressed in version 4.0. The DR4300e now supports 64 million files.

29576, 25733

Replication failure under some rare corner cases.

When upgrading from DR Series system software version 3.x to 4.0, the backup_user is migrated based on backup_user configuration in version 3.x. This is to minimize post-upgrade configuration.

29356

Configuration 1 - When RDA and OST are configured in 3.x,, the backup_user password is reset to default during the 4.0 upgrade.

Configuration 2 - When only RDA is configured in 3.x, the backup_user password is set to RDA during 4.0 upgrade.

Configuration 3 - When only OST is configured in 3.x, the backup_user password is set to OST during 4.0 upgrade.

29255

GUI: After upgrading the DR Series system software from 3.0.0.2 to 4.0, DR Series members are not listed in the Global View page.

Only Global View member lists from software version 3.1 and later will be migrated.

GV Members which were added in 3.0.x.x builds should be added/re-entered manually after upgrading the DR to 4.0.

28862

Time-out during seed copy to some slower USB devices.

Contact Technical Support to help you reduce the seeding threads so that the DR Series system does not overwhelm the client and USB device.

28561

Hyper-V client unable to discover VTL on higher-mapped LUN.

You must have a LUN 0-7 mapped to discover high-mapped LUN numbers. This is a Hyper-V issue.

28287

Unable to seed 3.x data to a 4.x system.

If you gathered a seed in 3.x you must finish your seed operation before upgrading to version 4.0.

28187

Setting a replication relationship with a target container has poor savings compared to the source.

If you create a replication relationship with a target container that already exists make sure the marker type is set to the correct type to match the source container.

27735, 27727

Deletion of replication pair with millions of files can take a very long time on target. This can also occur when reestablishing a relationship between a source and target container with a very large number of files.

Make sure replication is in sync before deleting the replication pair, or turn off replication before deleting the relationship. This will reduce the amount of unnecessary housekeeping needed if you are planning to delete the replication target.

27175

Upgrading from 1.x or 2.x to 4.0 is not prevented.

Your system should have 3.x before upgrading to 4.0. See the upgrade section at the beginning of these release notes.

26137

Under some conditions seeding job does not get removed from DR Series system when job is removed via CLI.

Unable to change the IP from Static to DHCP if using the same IP address.

25105

You must first enter a temporary IP address with a different number and then change to the IP you wish to use.

23767,21084

Disconnect under heavy load with Veeam, SQL, and other CIFS workflows.

Cleaner improvements in laundry processing size.

21715, 24242

This issue was addressed in the 3.2.0.2 maintenance release as well as version 4.0.

21496

On a DR2000v with ESX 5.5, web pages, pinging operations, etc. are timing out using TLB bonding.

Disable the second bond to work around this issue.

17216

The DR Series system is not Beast/Crime-vulnerable out of the box.

To keep this security intact, you should run one of the modern browsers that supports TLS 1.2 or greater.

16280

When you try to get replication statistics, the system displays the Software error: Use of uninitialized value in division (/) at helper.pl line 3548 error.

New Quest DLV license support.

30465

This feature was added in release 3.2.6.2 and 3.2.0.3 to support Quest and Dell license formats.

29733

Incorrect output was displayed when using the command cleaner --show_log_files.

This issue was addressed in release 3.2.6.2 and 3.2.0.3.

29618

This issue was addressed in release 3.2.6.2 and 3.2.0.3.

29540

Incorrect stats included in daily report for replication target.

This issue was addressed in release 3.2.6.2 and 3.2.0.3.

29421

Need to increase RPC client timeout for node start/stop.

This issue was addressed in release 3.2.6.2 and 3.2.0.3.

29266

Need to add cleaner tunable parameter for laundry size.

This feature was added to release 3.2.6.2 and 3.2.0.3. Note that this parameter must be tuned by Technical Support.

28499

Add cleaner GUI usage graph for cleaner details.

This feature was added to release 3.2.6.2 and 3.2.0.3.

28484

DR4300 systems displayed incorrect drive sizes/capacity for 4TB drives when using RM factor reset.

This was fixed in the release 3.2.6.1 RM package.

27625, 27555

Replication was encountering system crashes in some rare cases.

This was fixed in release 3.2.6.1.

26721

NDMP VTL device appears offline.

This was fixed in release 3.2.6.1.

26685, 25675

ocafsd was crashing on cleaner code in some rare instances.

This was fixed in release 3.2.6.1.

25908

NDMP restore fails when backup is restarted with CommVault Simpana with large amounts of data.

This was fixed in release 3.2.6.1, and NDMP Backup Restartable Extensions (BRE) are now supported.

28032

The DR4000 crashes when breaking a replication relationship and running fsck.

RDNFS plugin issue with Oracle Linux 7.0.

27847

Replication fix for large configurations.

27736, 27671

System cleaner timeout fix.

26191

After updating to release 3.2.0.1, the fan speed was 8K but should be 10K by default.

28439

This was fixed in release 3.2.0.2.

23879

Backup spanning across multiple tapes fail when 10 GB tape is used for Microsoft DPM.

Recommendations for backups spanning multiple tapes are documented in the DR Series best practices guide for Microsoft DPM available for your specific DR Series system at dell.com/support/home.

23641

Cannot join into the domain by adding a DR Series system to multiple OUs.

In the authenticate command type the format as follows:--ou topLevelOU/middleLevelOU/LowerLevelOU/TargetOU This format for adding OUs is now documented in the latest Dell DR Series System CLI Guide.

This was addressed and documented in release 3.2.6.

20870

RPC replication errors may be seen during replication from source DR Series system to target DR Series system. The primary and secondary replication target should have the same MTU setting. Replication proceeds once the MTU is set to the same as the replica target. This recommendation has been added to the section, “Managing Replication Operations,” in the latest Dell DR Series System Administrator’s Guide.

This was addressed and documented in release 3.2.6.

20539

After installing Rapid CIFS Filter Driver, an incorrect driver version is displayed.

Addressed and documented in release 3.2.6.

If at the end of successful installation, the version of the driver loaded (which can be checked using rdcifsctl.exe driver -q) is lower than the version of the installed package (listed under Control Panel > Installed products), then you should restart the machine. This issue can occur if the system fails to rename/delete driver files under the system folder.

20456

Installation of SSL Certificate file in the GUI failed on Internet Explorer (IE) when using Remote Desktop.

Addressed and documented in release 3.2.6.

When uploading the certificate file, the GUI uses the file name with the whole path, which works when uploading on a local system. When using RDP, in the IE browser settings, you need to disable the "Include local directory path .." setting, as follows.

Under Internet Options > Security, in the "Internet" or "Intranet" area, click Custom Level.

In the Security settings, scroll down to Include local directory path when uploading files to a server and ensure it is not selected.

Click OK and the refresh the browser.

20424

The following message may appear on the console: “vx_nospace...” even if the system still has space.

Addressed and documented in release 3.2.6.

This message is a result of trying to allocate a large contiguous block. The allocation of the contiguous block might fail, in which case, allocation continues without trying to keep the file contiguous. The vx_nospace error may appear, and this message can be safely ignored.

20188

When using cascaded replication for a large number of files (1M+), it may appear to take some time to see the actual files on the secondary target.

Addressed and documented in release 3.2.6. Namespace resync has to complete on the source and primary target and then data transfer has to start on both before transferring to the secondary target. Once the namespace resync completes on the first replication pair, resync then starts on the second replication pair. Until the namespace resync completes on the second pair, no data is transferred. This ensures the namespace resync completes faster and the snapshot can be deleted faster.

20009

User is able to create/write a file/directory in the share/container even though the user has only read permissions (file-level/security-level) on the share.

Addressed and documented in release 3.2.6. You can block write/create operations for a particular domain user by adding an extra ‘Deny’ ‘write’ acl for that user. (This issue occurs on the Windows 2008 client.)

19884

RDA plugin logging is not generated by default on physical hosts protected by vRanger because the plugin log service delldrlogsvc.exe is not installed by default.

Addressed and documented in release 3.2.6. To enable RDA plugin logging on physical hosts:

Copy %INSTALLDIR\Dell\RDA\dynlib\delldrlogsvc.exe from the vRanger server to the %INSTALLDIR\Dell\vRangerPhysicalClient\<timestamp>rofs directory on the physical host.

Change the directory to: %INSTALLDIR\Dell\vRangerPhysicalClient\<timestamp>rofs

Run the command: delldrlogsvc.exe install

Run the command: delldrlogsvc.exe start

19552

When there is no data to replicate (for example, replication state is INSYNC), replication statistics might indicate a target is running out of disk space even though the target has enough space.

Addressed and documented in release 3.2.6. This occurs because the replication process detected target space full condition when there is pending data to be replicated, but, at a later time, that pending data was removed and replication was insync. We do not query target disk state when replication is idle. On a similar note, we might not indicate target space full condition when there is no data to replicate as well. An error state in the replication statistics makes sense only when replication is not in sync.

19485

Adding a host name does not work if you use a FQDN and the host name has letters in it.

Addressed and documented in release 3.2.6. You can add names as long as the name is made only of numbers and it is not a FQDN where a period is used in the name string.

19186

If you change the hostname or the domain name after setting up the DR Rapid plug-in, then you may get the wrong backup mode pass-through or de-dupe.

Addressed and documented in release 3.2.6. Uninstall and then reinstall the DR Rapid plug-in for this information to be set correctly.

19023

Events information is not captured in the GUI when the date and time under System Configuration is changed to a date and time in the past.

Addressed and documented in release 3.2.6. This type of event information is captured in the CLI, by using the –show –alerts|more command.

18848

In the Internet Explorer 9 and 10 browsers, when you navigate to a DR Series system from within Global View, and then log out of that system and return to Global View, you are logged out of the machine running the Global View.

Addressed and documented in release 3.2.6. This issue does not occur with any other supported browser. For browser information, see the Dell DR Series System Interoperability Guide.

18178

The AIO-Stress tool fails when using Rapid NFS.

Addressed and documented in release 3.2.6. Rapid NFS, which uses FUSE, does not support AIO.

18501

Rapid NFS performance drops significantly with parallel NFS ingest from same client.

Addressed and documented in release 3.2.6. When using Rapid NFS on your client, it is suggested that you do not run NFS from the same client to the DR Series system in parallel as this will affect your overall performance.

18144

A File name too long error is displayed after trying to create a directory structure that exceeded 522 directories.

Addressed and documented in release 3.2.6. The limit on a path within a mount point is 4096 bytes.

18022

Windows 2008 and newer servers cannot assign local users (other than administrator) to Access Control Lists (ACLs). Windows 2008 and later does not query the local users on the DR Series platform.

Addressed and documented in release 3.2.6. Assign ACLs to the built-in Administrator user on the DR Series platform if using Windows 2008 and later.

17661

If the cable is removed from a network interface and the network is restarted, the interface does not get an IP. If you then reinsert the cable, you must restart the network to get an IP.

Addressed and documented in release 3.2.6. Ensure that the interfaces that you want to add or modify are connected when making Advanced Networking changes. To restart the network, type network --restart at the CLI command line.

17193

DR Rapid critical message Alert RDS Restore failed reported on DR4x00 system. In addition to the error, an event is generated internally in the DR Rapid plug-in when the read operation fails

Addressed and documented in release 3.2.6. The DR Rapid plug-in generates events for any kind of read failures.

16745, 17621

If you have the same client for OST and RDS operations and you manually update the client connections, unexpected messages are displayed as: [root@swsys-49 ~]# stats -- clients --type RDA RDA Client(s) Type plug-in OS Backup Software Last Access Connection(s) Mode r310-sys-39 RDS -- -- -- -- 0 Default

[root@swsys-49 ~]# ost -- update_client --name r310- sys-39 --mode auto Client is connected with default values. Please update to non-default values.

In Linux systems, the RDS plug-in installation fails if the OST plug-in is already installed in the system

Addressed and documented in release 3.2.6. To work around the issue, use a different client for OST and RDS rather than sharing the same machine for both protocols.

16660

When BE has an issue connecting to the server, it may not return detailed errors to the OST plug-in.

Addressed and documented in release 3.2.6. See the client log files for the detailed information on the connection failure.

16427

When upgrading your system using the iDRAC, if the upgrade package also contains new iDRAC firmware, the upgrade fails to complete.

Addressed and documented in release 3.2.6. Upgrade the DR code using the CLI or GUI using the normal network interface and not the iDRAC.

16053

After you install the Rapid CIFS plugin and replication operations are performed, you receive a blue screen and the following error: STOP 0x00000027 (parameter1, parameter2, parameter3, parameter4) RDR_FILE_SYSTEM

Addressed and documented in release 3.2.6. This is a Microsoft bug that occurs on shared network resources on a computer that is running Windows Server 2008 R2 or Windows 7. To fix, see the Microsoft KB2584874.

9433

The dedicated hot spare is not automatically configured after replacing the hot spare drive in slot 0.

Addressed and documented in release 3.2.6. If the hot spare is replaced, run maintenance by using the --make_standby command. Contact Dell Technical Support for this action.

27074

iSCSI VTL configuration with IBM Tivoli Storage Manager 7.1.1 fails at audit tape library.

Symantec Backup Exec inventory issue.

27074

Issue with an ofsck refcount record boundary calculation.

27013

NDMP VTL device offline issue.

26721

Kernel panic fix for VTL running in parallel with other protocol heavy replication.

26548

Option added to increase RDA retry capability in registry working with Tech Support.

26493

Support for RDA for Sparc.

26282

DR plugin DellDRLogSvc service does not always stop when uninstalling NetVault.

26265

Various VTL updates for release 3.2.6 ported to 3.2.0.1 code.

26190

RFE: OFSCK: re-organized refcnt record structure to optimize disk space.

26128

NFS fix for moujnt during stats call.

26113

Veeam ingest and replication issue for large file overwrite corner case.

25958

NetVault Backup stats file takes a long time in closing.

25860

ocafsd crashed on cleaner code due to corner case.

25675

Space handling improvements related to global space thresholds and per enclosure thresholds for Rapid NFS and Rapid CIFS.

25591

Improved seeding for CIFS on slow networks

25584

Cleaner is now set to run during idle time even when not in schedule. Since the machine is idle cleaning will occur; it will be stopped if I/O is started.

25393

Let cleaner finish if it is in runonce mode even if there is a cleaner schedule stop.

25135

cifs_iobuf_alloc fix for iSCSI.

25091

Fixed complex replication deadlock scenario involving overwrites and aborts.

24998

VTL subsystem starts if container name contains "ISCSI" or "NDMP" even though the container type is not 'vtl'.

24855

Corner case fix for cascaded replication.

24809

Validate the timestamps in mds supplied by user.

24738

DR VTL was showing support for hardware encryption, which is not accurate. Removed this option as Symantec was allowing encryption setting. (SPIN, SPOUT and publishing in MODE SENSE)

24701

Some of the documentation PDF links were not working in the DR Series system GUI.

24689

Unable to log in using domain user starting with letter "j".

24579

Enclosure showing as “Not Configured” when another enclosure is added, which changed ID numbers.

24571

Corner case where cleaner schedule will not stop cleaner from running, nor does the cleaner stop command.

23887

NDMP user name must be changed to ndmp_user in GUI.

23797