GUI client/REST server: *** Commit history: bdc24ab057..766a7bfab2 *** Auto deploy Web UI NG2 from Jenkins job sep-webui-ng2-520 build 123 (057505aa7f6108cdede73e77cb40adc02527b9f8) wip: Fix failing unit tests wip: (src#32135) Do not show parent migration tasks in active jobs view bugfix: (src#32135) Hide parent migration tasks via optional filter wip: (src#32135) Do not show parent migration tasks in active jobs view Auto deploy Web UI NG2 from Jenkins job sep-webui-ng2-520 build 122 (94f937c10854ffb8e9fc1c4719b0b8fc1ffb8f95) BUGFIX: (src#36293) sm_cmd dir ignores second tag in multi-tag filter wip: Fix failing unit tests BUGFIX: (src#36158) Remote directory on VMware vSphere datastore reported empty while it is not wip: (#36221) A scheduled archive adjustment does not operate in the mode specified by the media event BUGFIX: (#36221) A scheduled archive adjustment does not operate in the mode specified by the media event new: (src#36241) Backups/count fails when choosing savesetCntLessThan Auto deploy Web UI NG2 from Jenkins job sep-webui-ng2-520 build 121 (71b3c101bd724c4511d73d4df79401f1958deab9) wip: (src#36205) sm_cmd add taskgen doesn't work for autom. generation of PATH tasks and Windows root Filesystem BUGFIX: Generate PATH tasks for Windows root drives (src#36205) wip: (src#36186) Rework the "executeSql" endpoint to allow read-only access to the database only bugfix: Fix "countDynamic" statement to be executed when the "Last Backup State" view is shown. change: Set version name to "Artemis V6" Web restore assistant: *** Commit history: 1c01a389..57a8c291 *** wip: (src#36123) Nutanix datamover selection for restore according saveset format change: (src#36079) Just allow Linux hosts as data mover for VHS wip: (src#36123) Nutanix datamover selection for restore according saveset format Web client: *** Commit history: 2bbb0404c..057505aa7 *** wip: (src#36374) WEBUI: Migrations are not displayed under 'Active Tasks' BUGFIX: (src#36306) When creating a migration task, the D/I clicks from the WebUI are taken over incorrectly BUGFIX: (src#35922) In Dashboard the graph "Total restore data size over time" is empty Kernel: *** Commit history: 971c8130e6..cb83072b2c *** NEW: (#36029): allow postgresql-18 on debian trixie for installation on ubuntu 26.04 bugfix: (30984) Re-check pending jobs on global resource release change: Add overall throughput to sbc-3260 message wip: (#33223) Report current throughput by 'sbc' in human readable format bugfix: (src#36285) Stop SEP sesam BSR Pro asynchronously from the start of shutdown change: (src#36285) Add trace logging to stop_bsr_pro_service BUGFIX: (src#36285) Stop BSR Pro service after subprocesses stop bugfix: (src#36405) Fix successful backup/restore reported as ERROR BUGFIX: (#35049) If sbc_ndmp cannot resolve the storage host in combination with FTP, an application error occurs. Solution: End immediately change: (src#36405) Log halt_reason on questionable backup completion too change: (src#36405) Surface NDMP halt root cause at normal log level BUGFIX: (src#36405) Fix NDMP restore falsely reported as successful BUGFIX: (src#36395) Pass overall throughput to sbc-3260 in sesam_ndmp CHANGE: (src#36395) Log NDMP transfer size/throughput on a time basis change: Removed harmless sbc_ndmp warning for Isilon backup change: Removed ":" from the end of NDMP volume when displaying dir browse info bugfix: Correct for switch in 'sm_arch -O FORCE' bugfix: (#36397) 'Expire Media' overwrites tape, which just has been initialized shortly before BUGFIX: (src#36387) Fix DAR restore of tape-spanning NDMP saveset BUGFIX: (#34842) Allow restore of posix ACL's with invalid characters by using numeric ID's. Solution: Include the UID and GID during the backup. During restore, use the numeric UID/GID if available. Set the SESAM_ACL_NUMERIC environment variable to '0' to perform the restore by name. BUGFIX: (#36214) Single file restore of NDMP backup from tape failed. Solution: For FTP, the BCK_MESSAGE structure from the last successful login must be reused so that the IsTape property is preserved. bugfix: (#34353) Copy JSON datastore info file for all data stores. Reset 'iExists' variable in loop for every DS bugfix: (#36334) NDMP single item generation restore fails if folder from FULL and folder from INC are selected. Improve the logic for creating .ssl files for subtasks by checking whether the selected directories are also covered by other savesets bugfix: (#36334) NDMP single item generation restore fails if folder from FULL and folder from INC are selected. The sbc_ndmp must log sbc-3108item processed for every processed item, even if NDMP reports via 'Debug: Creating directory/file ...' bugfix: (#36334) NDMP single item generation restore fails if folder from FULL and folder from INC are selected. Solution: Match lines with backup_id BUGFIX: (#36334) NDMP single item generation restore fails if folder from FULL and folder from INC are selected. Solution: If only directories are selected then in case of a gen restore the lines must be fetched from the saveset's LIS file and used in the saveset specific SEL file. BUGFIX: (30984) sm_qm_main log gets too large. Problem: Log is flooded with lines such as 'Mode [idle] && 1 jobs are requesting resource single -> keep bJobRequest 1'. Solution: Avoid checking for pending jobs if nothing has happened. bugfix: (#36363) SEP Unit licence: activate the unreachable T2 'lent-to-T1' annotation in sm_info c (parity with Java -V 1) bugfix: (#34879) Exclude CD-ROM from backup (api v4) change: Nutanix api-v4 version before multi-stream backup NEW: (#30510) Add operating system 'Dell EMC OneFS' NEW: (#34879) Nutanix 7.5 - Prism Central support BUGFIX: (#36331) Single item generation restore of migrated backup fails due to wrong catalog file pattern. Solution: Generation restore sub task must use results.sesam_date to create path for original LIS bugfix: (#35931) Backup and restore complete /var/opt/sesam/var/lib/rear/etc/rear folder during rear update. Solution: Use 'cp' with '{path}/.' to assure proper copy despite of existing target directory change: (#34485) SEP sesam Tier Unit licence: use remaining upper level. Optimize output: Add amount of used upper or lower Units CHANGE: (src#34485) SEP sesam Tier Unit licence: use remaining upper level bugfix: (#35931) Backup and restore complete /var/opt/sesam/var/lib/rear/etc/rear folder during rear update. Use the backslash (\) before the command to prevent the alias from being expanded, here: alias cp='cp -i' bugfix: (#36243) Fix multiple boot_index 0 entries on VM restore BUGFIX: (#36273) fix GC slowness due to pageTrack growing size; release locked on restore pages after usage BUGFIX: (#35931) Backup and restore complete /var/opt/sesam/var/lib/rear/etc/rear folder during rear update BUGFIX: (#35921) Sesam Rear update on Server has to delete installation folder before extracting files. Solution: Use update routine from client rear update for server and client. BUGFIX: (#35225) Throughput and data size is not updated for external backups bugfix: (#35359) Hyper-V: Exit application in case of VM and Host incompability. Bugfix: Condition always true due to wrong operator '=' bugfix: (#36214) Single file restore of NDMP backup from tape failed. Bugfix: Commit f61c3be2 introduced regression with DISKHARD on Linux BUGFIX: (#36147) Just execute DB queries on SEP sesam server during shutdown of service. Skip DB queries on client, rds and gui installation. bugfix: erase all api-ms-win*.dll since Windows already has them built in. Revert "change: automatically added/removed files to signed file list." change: automatically added/removed files to signed file list. bugfix: (#36210) Tape drive encryption capability is switched off, if a non-encryptable medium is in the drive during configuration. Solution: Log sense ('-o') no longer printed the capacity of a loaded tape. BUGFIX: (#36210) Tape drive encryption capability is switched off, if a non-encryptable medium is in the drive during configuration. Solution: If no tape is loaded then avoid to print tape specific values: tape_remaining_capacity, tape_max_capacity BUGFIX: (#36210) Tape drive encryption capability is switched off, if a non-encryptable medium is in the drive during configuration. Solution: The command 'slu {device} -J short' now returns the correct value for 'tape_encryption_capability' even if the tape is not encryptable or no tape is loaded. Revert "BUGFIX: (#36210) Tape drive encryption capability is switched off, if a non-encryptable medium is in the drive during configuration. Solution: If no tape is loaded then avoid to print tape specific values: tape_remaining_capacity, tape_max_capacity" Change: (#35359) Hyper-V: Exit application in case of VM and Host incompability BUGFIX: (#36048) Only one MS-SQL restore can be performed at a time within the same instance. Add saveset to VDI device name change: (#27629) Redirected DB2 roll-forward fails, because new DB name is used in query. Change: Avoid misleading 'ERROR:' in trace message and the use of GLBV 'gv_conf_enforce_db2_source_match' NEW: (#36013) Do not submit backup jobs by 'sm_sepul_event', if Sesam shutdown is in progress BUGFIX: (#35048) Migration or external backup: The original error message is overwritten by 'E019-BACKUP Error processing the segment file:' bugfix: (#36021) Do not remove LIS files having '_SG_' in file name by 'sm_pur_status' bugfix: Drop trigger before re-creating called function bugfix: (#33953) Windows BSR backup fails with 'The handle is invalid'. Solution: Check exit state of 'sm_setup create_bootmedium_xml' bugfix: (#33953) Windows BSR backup fails with 'The handle is invalid'. Solution: Source 'all' uses 'sm_client dir /' and has to strip away " bugfix: (#33953) Windows BSR backup fails with 'The handle is invalid'. Solution: Do not attempt to print a BSR job script file that does not exist. BUGFIX: (#33953) Windows BSR backup fails with 'The handle is invalid'. Solution: Execute commands (sm_setup, sm_client) in try block to catch FileNotFound error NEW: (#36153) Add BSR Pro 22.2.260 to Sesam package. wip: (#36153) Add BSR Pro 22.2.260 to Sesam package. WIP: Add new keys to cm_hal_file_util.c sbc_bsr.py version/oo_msi_guids wip: (#36245) Add sleep in RDS alive check loop again wip: (#36245) Set correct number of queue slots wip: (#34625) Create drives queues and start sm_sms_watch, even if RDS isn't reachable. bugfix: (#36026) Create date based LIS file directory during migration of very old savesets BUGFIX: (#36180) Sesam startup fails, if 'sm_startup' runs with PID 11201. Correct 'regex' so PID is found only at the begin of line from 'ps' output BUGFIX: (#36181) Attempting to add a previously encrypted tape medium fails with the error message 'DB: ERROR: unterminated quoted string at or near ' Solution: Return the retrieved label in such case. BUGFIX: (#36148) If gv_ro_bin on RDS is not the same path as on Sesam Server sm_config_drives fails when executing sm_set_windows_defender_preferences.ps1. Solution: Do not prefix local gv_ro_bin NEW: (#33447) sm_cross_check: Write main log into 'gv_rw_prot_media' BUGFIX: (#36214) Single file restore of NDMP backup from tape failed. Solution: The following read operations were performed without setting the 'bIsTape' property. This resulted in an unnecessary byte-positioning request. BUGFIX: (#36211) Update DB:media.serial_number if not already set. Check for NULL and empty string BUGFIX: (#36210) Tape drive encryption capability is switched off, if a non-encryptable medium is in the drive during configuration. Solution: If no tape is loaded then avoid to print tape specific values: tape_remaining_capacity, tape_max_capacity BUGFIX: (#36274) Si3 status Si3 not written completely to Sesam DB. Increase internal buffer from 4096 to 6144 bytes, because status message could become longer now NEW: (#33573) Set termination time for 'sm_sbc_com_ext' job started by migration to same time set for 'sm_sms_copy' BUGFIX: (#36190) Replication job stucks in queue, if no source drive is set NEW: (#36179) Adapt dependenies for Sesam client on RHEL9 PPC new: Increase SEP sesam version. Old version is 5.2.0.24, new version is 5.2.0.25. new: (#36084) Multiple mount of same savese - fixed saving status to JSON for action prepare new: (#36084) Multiple mount of same saveset - save return starus of mount to JSON. If another mount for the same default mount dir is executed, return status from JSON is returned new: (#36084) Multiple mount of same saveset - saving mount info to JSON must not execute client_cleanup() if it is not for the real unmount of the last existing mount new: (#36084) Multiple mount of same saveset - Fixed get_info_value() on Windows. It returned the 1st value found for the same SSID because of \\ in paths new: (#36084) Multiple mount of same saveset - Formatted return message when mounting 2nd time for the same mount dir - 'Drives:' was missing new: (#36084) Multiple mount of same saveset - Implemented multiple mounts of VM backup types on Windows new: (#36084) Multiple mount of same saveset - close mapper file in VFS after VFS starts. Beacuse it was no closed, mapper file could not be replaced during second mount on different mount point and diskpart attach failed trying to mount non existing VHD, set from the previous mount's mapper uuid string bugfix: (#36084) Multiple mount of same saveset - umount was no exectuetd if info was noz found in json, but VFS was running new: (#36084) Multiple mount of same saveset - Action prepare saves mount info to json new: (#36084) Multiple mount of same saveset - Fixed unmount on Windows, mntdir path in JOSON has / as separator and / must be used when counting existing mounts for the same mount path new: (#36084) Multiple mount of same saveset - added additional logs change: (#36084) Increased timeout for semaphore for saving to mount in for json file to 10 seconds new: (#36084) Multiple mount of same saveset - implementation for VM backup types on Linux new: (#36084) Umount command without restore ID deletes all JSON entries for the given ssid after unmount of VFS mounts new: (#36084) Umount command without restore ID unmounts all Path mounts for the given saveset id bugfix: (#36084) Fixed typo error of the previous commit bugfix: (#36084) Removed function lock_info that was used only once and moved the locking of sema back to save_mounts_info(). This fixes the error of undefined variable sema new: (#36084) Multiple mount of same saveset - Implementation for backup type Path NEW: (#35821) Insert restart functionality in sm_sesam for Windows which is needed to restart SEP sesam service from WebUI. bugfix: (#36146) sm_replicate should write stack trace in Sesam main error log and trace file of sm_replicate BUGFIX: (#36081) If a Proxmox VM resides on LVM-thick, the snapshot cleanup reports 'No such file or directory' - Do not execute function to search for LVM-thin snapshots to be deleted if LVM resides on LVM-thick BUGFIX: (#35974) During a BSR backup, an exclude source of c:\temp\ is mapped to c/\temp\. Solution: Turn only the last char into '/' BUGFIX: (#35817) When performing a BSR backup, Sesam must specify the correct network interface for bootmedium.xml. Solution: Resolve the -c client name so that even a TCP/IP alias hostname is matched (by canonical name, alias list or resolved IP) and the correct IP address is determined. wip: (#27629) Redirected DB2 roll-forward fails, because new DB name is used in query. Wip: Add cm_lang.h wip: (#27629) Redirected DB2 roll-forward fails, because new DB name is used in query. Wip: sm_sbc_com: If tasks.exec = '0' then avoid backup bugfix: (#36094) ReaR must allow to use xorriso instead of ebiso. Solution: Use xorrisofs