BUGFIX: (#25497) Implement higher timeout of VDI connection for MSSQL operations. Solution: Increase timeout to 300 seconds
bugfix: (#29481) Incorrect formatted 'dir' for MSSQL module. Avoid leading '/'
BUGFIX: (#30241) The original error message is overwritten by 'No segment file {...}.dat found'. Solution: Overwrite backup message only in case of 'success' or 'with warnings' (recent commit 0c076ac7 was not correct)
BUGFIX: (#30613) Restore does not work, because sesam does not find the volume/directory in the lis file. Problem: The parent directory appeared before the recent volume. Solution: Push parent directory into delayed line list
BUGFIX: (#31450) Next Execution time is calculated incorrectly during winter to summer time change. Solution: Reduce one hour, except for a schedule that should start in the missing hour, e.g. at '2:30'. The hour - e.g. '2' - must be stored in next_exec to be used for the next calculation, e.g. add 2 hours to '2:30' results in '4:30'. READ_SINGLE_TERM() modified next_exec by calling tim_s2i()
BUGFIX: (#31781) SEP sesam Unit License. Problem: Backup type for 'HCL Domino' was not set to 'Lotus Notes' therefore the Unit Tier 3 was not correct.
BUGFIX: (#31821) MSSQL backup ends with (Arithmetic overflow error converting numeric to data type varchar). Solution: Use CAST(s.backup_size AS VARCHAR(20)) size in the sql statement.
bugfix: (#32209) Race condition reading 'sm.ini' on Windows. Do not use 'cm_fopen()' on Linux because it leads to many error messages searching correct 'sesam2000.ini' file
bugfix: (#32209) Race condition reading 'sm.ini' on Windows. Move cm_fopen() into common/common/cm_common.c
bugfix: (#32209) Race condition reading 'sm.ini' on Windows. Problem: x_Rename() must use MoveFile() if target file does not exist.
bugfix: (#32209) Race condition reading 'sm.ini' on Windows. Problem: x_Rename() must use MoveFile() if target file does not exist.
BUGFIX: (#32209) Race condition reading 'sm.ini' on Windows. Solution: Microsoft recommends using ReplaceFile() to rename files. New wrapper function cm_fopen() will retry fopen() if an error occurs. 'sm_setup set_java_path' without system_tray check.
BUGFIX: (#32256) Show all MS-SQL databases on one server by one 'dir' command. Problem: Prefix 'MS SQL Server:' must be removed from source for backup and restore.
bugfix (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: a large block did not fit in the buffer. Solution: refactored DB_WriteBlock function, added condition for setting DB_NUM_I_COMPLETE return value
BUGFIX (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: seeking in file every 1 mb. Solution: refactored logic of data reading
BUGFIX (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: the compiler sets the default size of the buffer of input streams to 4 KB. Solution: resizing the internal stream buffer according to the size of the block part
BUGFIX (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: the compiler sets the default size of the buffer of output streams to 4 KB. Solution: resizing the internal output stream buffer according to the size of the block part
bugfix (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: using the value of offset in file for every part of block, overflowing of payload blocks in some cases. Solution: added of offset inside the block and file offset, changed buffer size to 1 MB
bugfix (#32270): Hyper-V RCT INC/DIFF backup is much slower than FULL. Problem: very big buffer size for large blocks. Solution: set limit for raw pipe buffer size to 64MB
BUGFIX (#32270): Possible fix. Added cast of buffer type
BUGFIX (#32270): possible fix for restore issue. Problem: the buffer offset value overflowing. Solution: changed values from unsigned long to unsigned long long
BUGFIX: (#32297) Core dump sm_data_server and sm_stpd_http_conn after error allocating a memory. Solution: Check size before strcpy()
bugfix: (#32408) Backup of open sparse files without vss will not process all data. Compilation on Linux failed
BUGFIX: (#32408) Backup of open sparse files without vss will not process all data. Open with GENERIC_READ, FILE_SHARE_READ/WRITE to allow accessing a file which is opened by another process. Additional: Error handling improved.
bugfix: (#32411) Revert log on 'Unable to freeze filesystems' back to warning
bugfix: (#32411) variable was wrongy written and not defined. Changed warning 'Unable to freeze filesystems' to a normal log
bugfix: (#32432) Restore of big sparse file ~ 3GB will fail. Code revised. Even in case MTF_NO_SPAD the PDS_I_PROCESSED_FILE must be signalized.
BUGFIX: (#32443) INC Backup is not possible because of special F-D-I migration constellation. Problem: EOL of origin backup saveset was gone but saveset still exists. Solution: Do not retrieve already gone saveset use migrated saveset on other pool in that case.
bugfix: (#32455)  Migration of migration could fail with 'Invalid segment number',if 'DB:result_lbls' contains failed migration entry. Check for 'DB:result_lbls.saveset=0' in 'iFIND_MEDIUM()' function
BUGFIX: (#32486) Hyper-V RCT INCR or DIFF hung with 'Item enlarged during processing'. Problem: offset value overflowing. Solution: changed unsigned long to unsigned long long
BUGFIX: (32488) Sesam alarm interface script fails with NPE up from 5.1 when sending mail for a failed backup, because log file could not be found. Also use log file name for Windows set already by 'sm_backup' directly. For Linux use the given message from paramter $2 as mail body and add the log file as attachment instead of putting it in the mail body.
BUGFIX: (#32488) Sesam alarm interface script fails with NPE up from 5.1 when sending mail for a failed backup, because log file could not be found. Use log file name set already by 'sm_backup' directly
bugfix: (#32512) Restored VMs with new name have the same GUID as the original VM. Problem: restor to original generate new Bios GUID. Solution: added condition for generating new GUIDs
bugfix: (#32512) Restored VMs with new name have the same GUID as the original VM. Problem: using original VM configuration. Solution: added script which will change VM Bios GUID after successful importing
bugfix: (#32520) Error during calculation of next starting time for 1 hourly schedule. Solution: The calculation allowed only 3600 loops, so an hourly schedule advanced only 150 days. Increased to 100000 (11 years)
BUGFIX: (#32526) MSSQL incr log backup ends with unequal lsn number on a database. Solution: Improve LSN check: Log backup must not take care about previous Diff backups, only check if previous basic incr saveset last_lsn == first_lsn, if previous backup was a full then first_lsn <= full last_lsn+1 <= last_lsn and in case of Diff check database_backup_lsn == Full checkpoint_lsn
BUGFIX: (#32528) Update date in 'sm.ini' not set correctly during Linux Sesam server update. Fix parsing of '' token in parser
bugfix: (#32531) Recovery level for database not correctly detected if backup source contains server and instance name. Solution: If full qualified database name server/[instance]/database then select status from master..sysdatabases for given database.
BUGFIX: (#32534) Wrong relocation parameter for VMDK restore set. Problem: Special symbols in relocation have to be masked with back slash as a result string size enlarged. Solution: Correct string size calculation
BUGFIX: (#32543) Upload of vSphere '.nvram' fails, if datacenter contains UTF-8 character. URL encode datacenter name set by UI too
bugfix: (#32545): Revert all changes made in commit '668156f5b9df' in sm_config_drives merged from 'develop' branch
bugfix: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Problem: Commit 6825e0bd introduced abort after catching first signal, e.g. in sm_restore. Solution: Add macros to improve signal handling in sm_sms_backup, sm_restore, sm_sms_copy. At least 2 signals should be caught.
bugfix: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Problem: Recent merge conflict resolution lead to code lines of develop branch which are not yet required in sm_config_drives
bugfix: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Problem: Revert recent change in sm_drive
bugfix: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Problem: Windows: Compilation failed sesam.def : error LNK2001: unresolved external symbol gcInterrupt
BUGFIX: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Solution: Now an additional RCMD flag indicates that the caller wants an immediate termination when the signal is caught. Implemented for sm_sms_backup, sm_restore, sm_sms_copy.
BUGFIX: (#32545) sm_alarm interface couldn't be executed, if backup has been aborted. Solution: Reset interrupt flag
BUGFIX: (#32557) Fix setting of image transfer timeout via parameter
BUGFIX: (#32557) OLVM/RHEV backup fails due to imageservice timeout: set higher timeout value than default of 60 seconds, allow setting timeout via arguments
BUGFIX: (#32557): VM is locked in backup state if transfer ticket expired: wrap setting up connection to transfer service in try/catch case so error can be catched and backup state is finalized
bugfix: (#32570) sbc cores during restore to a different Nutanix cluster. Init variable
BUGFIX: (#32578) Sesam service doesn't start, if STPD INI files couldn't be read by Python config parser. Ignore exception and continue with next file
BUGFIX: (#32597) Migration does not work for  savesets with DB:result_lbls.saveset_state='1'
bugfix: (#32601) A large Log LSN number (>64Bit) leads to a wrong result when comparing the Log first LSN with the Full last LSN. Problem: Build failed for sbc_com_interface due to different itoa() function type. Problem with argument length in csae of ARG_SYSTEM also fixed.
bugfix: (#32601) A large Log LSN number (>64Bit) leads to a wrong result when comparing the Log first LSN with the Full last LSN. Problem: Linux does not provide itoa() so use string_arithmetic.c szSTR_itoa( int n, char s[] )
bugfix: (#32601) A large Log LSN number (>64Bit) leads to a wrong result when comparing the Log first LSN with the Full last LSN. Solution: Build failed due to several definitions of NUM_LEN
BUGFIX: (#32601) A large Log LSN number (>64Bit) leads to a wrong result when comparing the Log first LSN with the Full last LSN. Solution: Increase NUM_LEN to avoid overflow in string arithmetic
BUGFIX: (#32601) A large Log LSN number (>64Bit) leads to a wrong result when comparing the Log first LSN with the Full last LSN. Solution: Use string arithmetic.
BUGFIX: (#32605) Hyper-V RCT backup accidentally destroys all user snsapshots. Problem: call the function for destroing user snapshots. Solution: delete the call of the function
bugfix: (#32613) Backup of a single file over EOM fails using LISv2. Solution: Break while loop when 'Total: ...' line is reached
BUGFIX: (#32615) enable RCT backup with configuration version 10
BUGFIX: (#32629) vSphere SFR via cross mount fails with server on Windows and RDS on Linux. Fixed path separator'
BUGFIX: (#32648) MSP license must not add BSR licenses to Level 1. Solution: Stop processing when END_OF_SESAM_BACKUP_TYPES is reached.
bugfix: Sles11 does not know 'retricted'
bugix: (#30250) SQL statement error: ERROR: more than one row returned by a subquery used as an expression. Solution: Adaption for select statment was missing
change: (#22358) MS SQL Server Recover with Point_in_Time and Recovery state option. Change: Only Diff backup must check for current database_backup_lsn == recent Full checkpoint_lsn. Log (INCR) must only check for last_lsn == first_lsn
change: (#28353) Continue selective restore then ESET is missing. Restore should be marked as "successful with warnings"
CHANGE: (#32537) Change Java dependency for Sesam server on SLES12 from 'jre >= 11' to 'java-11-openjdk', because recent Java RPMs from SuSE doesn't provide 'jre >= 11' tag any more
NEW: (#31049) Enhance backup chain check for MSSQL log backup. Solution: Use the basic saveset which is given in the saveset info to retrieve previous backed up log last_lsn number
new: (#32256) Show all MS-SQL databases on one server by one 'dir' command
new: (#32256) Show all MS-SQL databases on one server by one 'dir' command
new: (#32256) 'sm_sho dir -o recursive <client> <path>' executes 'sm_client dir -r ...' on client
NEW: (#32415) Windows backup fails with 'Not supported Microsoft reparse tag 0x80000023'. Solution: Add support for IO_REPARSE_TAG_AF_UNIX - Linux (WSL) UNIX domain socket
NEW: (#32467) Windows backup fails with 'Not supported Microsoft reparse tag 0x00000036' (IO_REPARSE_TAG_CITRIX_PM). Solution: Add support for IO_REPARSE_TAG_CITRIX_PM - Citrix Systems for profile management
NEW: (#32488) Set correct backup log path (NOT log) for backups, which terminate in pending queue
new: add version.h for hyperv_rct module
new: Increase SEP sesam version. Old version is 5.1.0.7, new version is 5.1.0.8.
wip: (#26035) Copy Sesam kits from Jenkins artifactory into download dfolder structure.
wip: (#32487) - MS-SQL task generation, read all databases from all instances and create tasks