Release of Service Pack 1 of Beefalo V2: Difference between revisions

From SEPsesam
(Draft for SP1 Beefalo V2 release (proposed by EE), in progress.)
 
(Minor correction.)
(11 intermediate revisions by 2 users not shown)
Line 1: Line 1:
{{Draft}}
__NOTOC__
Service Pack 1 of [[SEP_sesam_Release_Versions|''Beefalo V2'']] (4.4.3.84 SP1) is now available. It is highly recommended to install SP1 as it includes various performance, stability and key fixes. It can be downloaded from {{download}}. For the latest release, see [[Release_Notes_4.4.3_Beefalo_V2|''Beefalo V2'' Release Notes]].
Service Pack 1 of [[SEP_sesam_Release_Versions|''Beefalo V2'']] (4.4.3.84 SP1) is now available. It is highly recommended to install SP1 as it includes various performance, stability, and other key customer fixes not available in the [[Release_Notes_4.4.3_Beefalo_V2|initial release of ''Beefalo V2'']].  


== Fixed known issues and limitations ==
* Service Pack SP 1 (1716dc3) of ''Beefalo V2'' (4.4.3.84 SP1) - released: 14<sup>th</sup> of October, 2020. See [[SEP_sesam_Release_Versions|SEP sesam Release Versions]]. 
 
[[File:SEP Tip.png|45px|link=https://download.sep.de]][https://download.sep.de/servicepacks/4.4.3/4.4.3.84/ Download SP1]
 
== Known issues ==
====Hyper-V restore of a virtual machine to a different location with a new name may overwrite the original VM====
'''''Problem'''''
 
With version ''Beefalo V2 SP1'' (4.4.3.84 SP1), Hyper-V restore of a virtual machine to a different location with a new name may overwrite the original VM, resulting in complete loss of data stored on the original VM. This only happens if the original VM is offline (powered off) at the time of a restore. This issue only applies if you are running ''Beefalo V2 SP1''. Hyper-V restore in the previous versions is working fine.
 
'''''Workaround'''''
 
Do not perform a Hyper-V VM relocation restore with SEP sesam ''Beefalo V2 SP1''. This issue is already fixed and an official fix will be released soon. We will post a notice as soon as the new service pack is available.
 
== Fixed known issues ==


=== GUI ===
=== GUI ===
*VM restore with vCenter 7 fails with the API version error, i.e., ''Version incompatible: apiversion '7.0.0' of vCenter''.
====VM restore with vCenter 7 fails====
*Archive adjustment ''by barcode only'' does not work.
 
*If a system load is very high, the <ttt>sm_cmd</tt> may hang and blocks other processes.
'''''Problem'''''
*Relocation of MS SQL database via GUI does not work, because it does not provide logical names.
 
*Single file restore from a Hyper-V (''COPY'') backup fails.
VM restore with vCenter 7 fails with the API version error, i.e., ''Version incompatible: apiversion '7.0.0' of vCenter''.
 
'''''Cause'''''
 
Enhancement: Support for VMware vSphere 7.0 (ESXi and vCenter) .
 
====Archive adjustment ''by barcode only'' does not work====
 
'''''Problem'''''
 
The archive adjustment with the option ''Adjustment by barcode only'' does not work.
 
'''''Cause'''''
 
The value of the <tt>-n</tt> parameter was missing.  
 
====The <tt>sm_cmd</tt> hangs====
 
'''''Problem'''''
 
The <tt>sm_cmd</tt> hangs and blocks other processes.
 
'''''Cause'''''
 
If a system load was very high, the <tt>sm_cmd</tt> may hung and block other processes.
 
====MS SQL relocation does not work====
'''''Problem'''''
 
The relocation of the MS SQL database via GUI does not work.
 
'''''Cause'''''
 
The MS SQL relocation did not provide logical names.
 
====Hyper-V single file restore fails====
'''''Problem'''''
 
Restoring a single file from a Hyper-V (''COPY'') backup fails.
 
'''''Cause'''''
 
The restore wizard drives combo box was empty.


=== Kernel ===
=== Kernel ===


*Remote command hangs, if one line is longer than 16382 bytes.
====Remote command hangs if one line is too long====
*On some Linux systems a backup to tape fails because of unsupported read block size - ''errno EBUSY''.  
'''''Problem'''''
*Mounting a VM with multiple disks on Linux mounts only one disk or fails to mount completely.
 
*vSphere restore fails with ''vmx path has no extension'' because the VM information is cut off.  
Remote command hangs if a line is longer than 16382 bytes.
*The obsolete GLBV are not deleted.  
 
*VM backup fails, if a VM name consists only of numbers (especially on Proxmox VE and OpenNebula)
====Backup to tape fails on Linux====
*Restore fails if the SMS data server does not reset a block size to variable after write operation with LTO 7/LTO 8.  
'''''Problem'''''
*Selective restore of a Windows saveset may restore wrong files or it may fail with MTF error due to ESET after a new segment (mainly on the RDX drives).
 
*The <tt>sm_config_hw</tt> changes the device name of tape drive to the wrong value.  
On some Linux systems, a backup to tape fails because of unsupported read block size - ''errno EBUSY''.  
*The savesets from failed migration are used for VMDK mount operation.
 
*EOM during restore; when using a tape media pool, during the tape change (''Write'' operation) the next label is not known and the restore fails.  
====Mounting a VM with multiple disks on Linux====
*''SESAM_BACKUP'' on Windows ends with warning due to thread exit code <tt>STATUS_THREAD_IS_TERMINATING</tt> with disaster interface.
'''''Problem'''''
*Hyper-V single file restore from RDS fails with ''The system cannot find the path specified''.
 
*Restore from the RDX disk drive stalled, STPD log shows an endless loop.
Mounting a VM with multiple disks on Linux mounts only one disk or fails to mount completely.
*Concurrent processes created same unique ID.
 
*MS SQL backup is not possible if ''TLS 1.0'', ''SSL 3.0'' and ''SSL 2.0'' are disabled.
====vSphere restore fails with ''vmx path has no extension''====
*Update SQL statement for ''DB:results.source'' fails, because the size of SQL command string is too long.
'''''Problem'''''
 
vSphere restore fails with ''vmx path has no extension'' when the VM name is too long.
 
'''''Cause'''''
 
The VM information was cut off.  
 
====Deleting obsolete GLBV====
'''''Problem'''''
 
The obsolete global variables (GLBV) are not deleted.  
 
====VM backup with integer VM name fails====
'''''Problem'''''
 
VM backup with integer VM name fails with ''../sbccom.py", line 134, in closejob: cannot concatenate str and int objects'' (especially on Proxmox VE and OpenNebula).
 
'''''Cause'''''
 
The problem occurred with the backup sources (VM names) that consisted of numbers only.
 
====Restore read size issue with LTO 7/LTO 8====
'''''Problem'''''
 
The restore fails if the SMS data server does not reset a block size to the variable block size after write operation with LTO 7/LTO 8.  
 
'''''Cause'''''
 
The restore read size is incorrect. Consequently, 2 blocks are read in one read operation. Setting a variable block size for LTO drives resolves the problem.
 
====Selective restore on Windows restores wrong files or fails with the MTF error====
'''''Problem'''''
 
Selective restore of a Windows saveset may restore wrong files or it may fail with MTF error due to ESET after a new segment (mainly on the RDX drives).
 
'''''Cause'''''
 
SSET generated false offset2 for ESET.
 
====<tt>sm_config_hw</tt> changes the device name====
'''''Problem'''''
 
The <tt>sm_config_hw</tt> changes the device name of tape drive to a wrong value.  
 
====VMDK mount operation====
'''''Problem'''''
 
VMDK mount operation uses the savesets from failed migration.
 
====EOM during restore====
'''''Problem'''''
 
When using a tape media pool, during the tape change (''Write'' operation) the next label is not known and the restore fails.  
 
====''SESAM_BACKUP'' on Windows ends with warning====
'''''Problem'''''
 
''SESAM_BACKUP'' on Windows ends with warning due to thread exit code <tt>STATUS_THREAD_IS_TERMINATING</tt> with disaster interface.
 
'''''Cause'''''
 
Thread exit code <tt>STATUS_THREAD_IS_TERMINATING</tt> is not handled properly.
 
====Hyper-V single file restore from RDS fails====
'''''Problem'''''
 
Hyper-V single file restore from RDS fails with the error ''The system cannot find the path specified''.
 
'''''Cause'''''
 
Hyper-V backup with an incompatible file system was used for mounting.
 
====Restore from the RDX disk drive stalled====
'''''Problem'''''
 
Restore from the RDX disk drive stalled and the STPD log generates log files in an endless loop.
 
====Concurrent processes created same unique ID====
'''''Problem'''''
Concurrent processes create the same unique ID. Failure during access to database: SQL query. Compilation with Visual Studio 2019 failed due to unresolved time zone and daylight.
 
====MS SQL backup via ODBC fails====
'''''Problem'''''
 
MS SQL backup via ODBC fails with the protocol error.
 
'''''Cause'''''
 
MS SQL backup is not possible, if ''TLS 1.0'', ''SSL 3.0'' and ''SSL 2.0'' are disabled. The problem was the obsolete ODBC SQL Server driver.
 
====Update statement of results fails when the source is too long====
'''''Problem'''''
 
Update SQL statement for ''DB:results.source'' fails.
 
'''''Cause'''''
 
The size of the SQL command string is too long. Resolved with the correct handling of SQL statements using a backup source longer than 1024 or 2048 characters.

Revision as of 12:50, 13 November 2020

Service Pack 1 of Beefalo V2 (4.4.3.84 SP1) is now available. It is highly recommended to install SP1 as it includes various performance, stability, and other key customer fixes not available in the initial release of Beefalo V2.

SEP Tip.pngDownload SP1

Known issues

Hyper-V restore of a virtual machine to a different location with a new name may overwrite the original VM

Problem

With version Beefalo V2 SP1 (4.4.3.84 SP1), Hyper-V restore of a virtual machine to a different location with a new name may overwrite the original VM, resulting in complete loss of data stored on the original VM. This only happens if the original VM is offline (powered off) at the time of a restore. This issue only applies if you are running Beefalo V2 SP1. Hyper-V restore in the previous versions is working fine.

Workaround

Do not perform a Hyper-V VM relocation restore with SEP sesam Beefalo V2 SP1. This issue is already fixed and an official fix will be released soon. We will post a notice as soon as the new service pack is available.

Fixed known issues

GUI

VM restore with vCenter 7 fails

Problem

VM restore with vCenter 7 fails with the API version error, i.e., Version incompatible: apiversion '7.0.0' of vCenter.

Cause

Enhancement: Support for VMware vSphere 7.0 (ESXi and vCenter) .

Archive adjustment by barcode only does not work

Problem

The archive adjustment with the option Adjustment by barcode only does not work.

Cause

The value of the -n parameter was missing.

The sm_cmd hangs

Problem

The sm_cmd hangs and blocks other processes.

Cause

If a system load was very high, the sm_cmd may hung and block other processes.

MS SQL relocation does not work

Problem

The relocation of the MS SQL database via GUI does not work.

Cause

The MS SQL relocation did not provide logical names.

Hyper-V single file restore fails

Problem

Restoring a single file from a Hyper-V (COPY) backup fails.

Cause

The restore wizard drives combo box was empty.

Kernel

Remote command hangs if one line is too long

Problem

Remote command hangs if a line is longer than 16382 bytes.

Backup to tape fails on Linux

Problem

On some Linux systems, a backup to tape fails because of unsupported read block size - errno EBUSY.

Mounting a VM with multiple disks on Linux

Problem

Mounting a VM with multiple disks on Linux mounts only one disk or fails to mount completely.

vSphere restore fails with vmx path has no extension

Problem

vSphere restore fails with vmx path has no extension when the VM name is too long.

Cause

The VM information was cut off.

Deleting obsolete GLBV

Problem

The obsolete global variables (GLBV) are not deleted.

VM backup with integer VM name fails

Problem

VM backup with integer VM name fails with ../sbccom.py", line 134, in closejob: cannot concatenate str and int objects (especially on Proxmox VE and OpenNebula).

Cause

The problem occurred with the backup sources (VM names) that consisted of numbers only.

Restore read size issue with LTO 7/LTO 8

Problem

The restore fails if the SMS data server does not reset a block size to the variable block size after write operation with LTO 7/LTO 8.

Cause

The restore read size is incorrect. Consequently, 2 blocks are read in one read operation. Setting a variable block size for LTO drives resolves the problem.

Selective restore on Windows restores wrong files or fails with the MTF error

Problem

Selective restore of a Windows saveset may restore wrong files or it may fail with MTF error due to ESET after a new segment (mainly on the RDX drives).

Cause

SSET generated false offset2 for ESET.

sm_config_hw changes the device name

Problem

The sm_config_hw changes the device name of tape drive to a wrong value.

VMDK mount operation

Problem

VMDK mount operation uses the savesets from failed migration.

EOM during restore

Problem

When using a tape media pool, during the tape change (Write operation) the next label is not known and the restore fails.

SESAM_BACKUP on Windows ends with warning

Problem

SESAM_BACKUP on Windows ends with warning due to thread exit code STATUS_THREAD_IS_TERMINATING with disaster interface.

Cause

Thread exit code STATUS_THREAD_IS_TERMINATING is not handled properly.

Hyper-V single file restore from RDS fails

Problem

Hyper-V single file restore from RDS fails with the error The system cannot find the path specified.

Cause

Hyper-V backup with an incompatible file system was used for mounting.

Restore from the RDX disk drive stalled

Problem

Restore from the RDX disk drive stalled and the STPD log generates log files in an endless loop.

Concurrent processes created same unique ID

Problem Concurrent processes create the same unique ID. Failure during access to database: SQL query. Compilation with Visual Studio 2019 failed due to unresolved time zone and daylight.

MS SQL backup via ODBC fails

Problem

MS SQL backup via ODBC fails with the protocol error.

Cause

MS SQL backup is not possible, if TLS 1.0, SSL 3.0 and SSL 2.0 are disabled. The problem was the obsolete ODBC SQL Server driver.

Update statement of results fails when the source is too long

Problem

Update SQL statement for DB:results.source fails.

Cause

The size of the SQL command string is too long. Resolved with the correct handling of SQL statements using a backup source longer than 1024 or 2048 characters.