Releases: mantechnology/bsr
1.6.4.0 Patch Release
Release Date
May 26, 2023
Issues
Improvement
- Improve reconnection due to delayed ping response at start of sync
Bug
- Fixed a command hang on the DR node during the 1:2 iteration of secondary node down/up during replication.
1.6.3.0 Full Release
Release Date
Mar 31, 2023
Issues
New Feature
- Support for Mixed Windows/Linux Configurations
- Supports Ubuntu 22.04
- Support for Rocky 8.7, 9
- Added service operation log
- Save bsr kernel log files
- Added bsrmon IO pending monitoring entry
Change
- Change property disable_write_same to enable by default
- Do not save duplicate logs and status inquiry logs among CLI logs
- Remove [windows] preshutdown, ServiceStart, ServiceStop logs
Story
- Support for low-end VM environments
Bug
- Target node disk status is updated after disconnection with bsrsetup disconnect command
- In 1:2 connections, the state of the other node remains standalone during connect/disconnect iteration
- In a 1:2 connection, when demoted during congestion, the node is not synchronized and remains behind after congestion is resolved.
- Unreleased listener reference panic when running up/down concurrently in 1:2+ configuration
- Deadlock when executing bsr_adm_new_path() while bsr_adm_down() is executing
- Split brain when connecting to old source node during synchronization in 1:2 configuration if another node is promoted (younger primary) after terminating connection with source node
- [Linux] Synchronization does not proceed with Sync_Source / Behind state while automation is in progress
- Panic with synchronization timer registration after connection termination
- Abnormal termination when doing bsrmon /watch resync_ratio in multivolume configuration
- Split brain occurs in 1:2 configuration if node's state is different when sending and receiving UUIDs
- Detect oos when replication occurs in the verify in-progress zone
- SyncSource remains 100% when secondary node down/up during IO occurrence
- Replication, target node left out of sync after reconnection during synchronization
- Target node remains out of sync with other target nodes whose uuid was received from target node during synchronization
- Replication, synchronization does not start on source node after reconnection during synchronization
- Replication, remains out of sync when twopc abort for connection occurs after reconnection during synchronization
- Running del-path command while unconnected, subsequent reconnections consistently fail
- Replication, synchronization not progressing on reconnection during handler call during synchronization
- Synchronization with promotion, no sync zone and uuid update during synchronization leaves out of sync on subsequent target nodes
- Synchronization completion does not update to UUID received during synchronization completion
- When using interface index in ipv6 configuration, execution of adjust causes del-path to be performed
- After synchronization completion, history UUID of another node is set to current UUID, leaving out of sync
- Out of sync sent by the source node is received by the target node after synchronization is completed, leaving out of sync.
- [windows] listener removal fails to return lock due to wait during removal, resulting in a system line
- Performance data collected per peer in ipv6 floating configuration is not recorded
- [bsod] peer_device reference removed due to connection object being removed and out of sync in unacked peer_request sent
- [rhel 9.1] Panic in QUEUE_FLAG_DISCARD flag removal logic when running resource up
- [bsod] Use of mutex in spinlock area on device removal
- Synchronization does not proceed after congestion condition
- [BSOD] uuid rotate occurs when no peer_device object exists
- [BSOD] BSOD on primary node with both node disks offline while replication is in progress
- [BSOD] Occurs when setting oos on an invalid peer_device
- Error in bsrmon /show output on ipv6 floating configuration
- [linux] secondary node write IO error, kernel panic when changing to diskless
- [Windows] No system logs collected when performing bsrsupport after bsrsupport -exclude_systemlog system log exclusion collection
- Target node does not continue to synchronize with other nodes when source node poweroff occurs during synchronization in 1:2 configuration
- [Windows] Syntax error when performing bsradm command after ipv6 link local configuration
- Async + sndbuf 0 environment takes 900 seconds to disconnect due to timed_out
- Synchronization does not proceed on target node outdated/behind during asynchronous environment file replication.
- [Linux] OS panic when sndbuff-size value is unsigned
Improvement
- Performance Monitor query range improvement
- Improve bsrmon output
- Provide bsrmon json output option
- Add log for UUID analysis
- Provide command to set synchronization method (disable FastSync)
- Provide option to exclude system log collection when collecting support files
- Remove dummy oos generated in disconnected state
- Improved error handling for data transfers
- [windows] Support setting scopeid interface alias when using ipv6 link-local
- Improve configuration method when both nodes have different IPv6 Scope IDs
- [windows] Fix inf validation error when signing driver hardware
1.6.2.4
Release Date
DEC 22, 2022
Issues
Change
- Rename the windows bsr service (bsrservice -> bsr)
Improvement
- Remove existing loaded modules when installing a linux kernel module rpm upgrade
- Improved network retransmission timeout detection
- Improved behavior on error in transit (-EAGAIN)
Bug
- System BUG log "Objects remaining in bsr_req" when unloading ubuntu 20.04.2 module
- Synchronization does not proceed to Consistent state on target node when executing synchronization command
- Synchronization not performed on initial promotion during connection, both nodes have different roles
- Kernel panic due to deleted peer_device reference in debugfs during down
- Panic on bsr_get_connection_by_node_id+0x9/0x50 NULL reference
- System hang due to device->timing_lock deadlock
- [LINUX] Panic terminating connection with one node in a 1:2 or larger connection
- fixed bsrmon memory leak
- Fixed addr_equal+0x5/0x70 panic during connection establishment
- [LINUX] Kernel panic when executing a down command before completing a bsradm up command
- Inconsistency due to write failures when performing UP after resource initialization
1.6.2
Release Date
Aug 18, 2022
Issues
New Feature
- Highwater congestion method, congestion detection mechanism based on the number of packets
- Rocky Linux, Windows 11/2022 support
Improvement
- Add events for node name change
- Improve minimum sync-rate, Guaranteed minimum synchronization bandwidth.
Bug
- Fixed an issue that socket data transmission was not retransmitted if missing
- Fixed stack dump error problem when host method of host section and connection section does not match
- Fixed disconnection issue with (BAD! BarrierAck) on multiple I/O in asynchronous configuration
- [Linux] Fixed an issue where resources were not automatically launched when service stop -> start in multi-resource configuration
- Fixed BSOD issue in down operation sometimes when performing secondary down -> up during asynchronous replication
- Fixed the problem that the uuid of the source was not updated due to disconnection after synchronization was completed
1.6.1
Release Date
Mar 25, 2022
Issues
New Feature
- prolinux support
- HVCI support
- CLI log
- new bsrmon feature
- add bsrfsflt bypass control interface
Change
- Do not unload kernel module on linux service shutdown - prevent the events2 command from exiting
Improvement
- Improved connection speed when configuring MDR-DRX (internal)
- Port redundancy support for each connection in DRX interworking configuration
- Add activity log performance monitor entry
- Added command to check performance data aggregation status
- Send buffering status monitoring
Bug
- inconsistency in replication and synchronization in ahead status
- inconsistency due to IO completion order change when replication and synchronization are in progress in ubuntu 18.04.06
- the issue where svc_auto_up is not started even though svc_auto_up is yes in more than one resource
- PAGE_FAULT_IN_NONPAGED_AREA BSOD occurs when disconnect is performed after verifying
- Timeout issue when executing synchronization command in large disk environment
- bsradm status command invalid host error when nic is disabled in floating configuration
- The case that the connection is not connected and the primary --force fails in a large-scale replication environment
- A case in which before-resync-target works even though there is no handler attribute value and exit code 66 is output.
- Inconsistency during disk removal of both nodes
- A case in which NetworkFailure status is maintained when network is disconnected during replication
- Full sync operation due to failure to allocate bitmap buffer for fast sync during initial synchronization
- PAGE_FAULT_IN_NONPAGED_AREA BSOD occurs in target node during file copy + synchronization
- Down failure during MDR resource batch down command through replix
- A case that the status is stopped when verify command after SDR online
- A case in which the synchronization completion log is output twice in secondary when synchronization is in progress following promotion to primary
- GI change event does not occur when bitmap uuid is removed
- When special IO occurs during synchronization, oos remains in primary after synchronization is completed
- svc-auto-up no; Cloned disk is not locked when rebooting after setting
- BSOD in del-path during disconnecting
1.6
Release Date
Aug 12, 2021
Issues
New Feature
BSR-659 Linux Secure Boot Support
BSR-593 Add svc-auto-down option
BSR-511 Provides Windows Safe Mode Recovery
BSR-457 xfs FastSync
BSR-450 ext FastSync
BSR-219 oos trace log
BSR-118 FastOV support
BSR-114 Log watch support
BSR-12 Porting Linux Send Buffer Support
BSR-5 Loop meta support
Change
BSR-718 Move svc-auto-up and svc-auto-down options to node section
BSR-458 Change the policy to bypass on the engine for volume read I/O
BSR-124 Changing GUID Notation When Configuring Windows Resources
Bug
BSR-731 [Linux] Fixed an intermittent crash when removing the hard disk during cloning
BSR-708 Fixed the problem that GI information was not passed as an event
BSR-674 [BSOD] Fixed BSOD issue when installing WDRBD and BSR together
BSR-617 Fixed the problem that the disk size is not updated when executing the bsrcon /info command after reducing the disk
BSR-514 Fix missing bsrcon /info output size, guid
Improvement
BSR-766 [Linux] Added P_ZEROES packet for processing REQ_OP_WRITE_ZEROES bio
BSR-676 Event notification when GI changes
BSR-671 Improved windows bsr log file output
BSR-664 Add disk information when collecting support files
BSR-626 Port handler_use to Linux.
BSR-624 Improved output when rolling logs while running bsrwatch
BSR-127 Improved verify performance through transmit buffering in protocol C
BSR-122 Save real-time log files and support log rolling
BSR-109 Improved Windows mount information update method (IOCTL_MOUNTDEV_LINK_CREATED)
1.6-A13
Release Date
Jul 6, 2021
Issues
Bug
[BSR-768] Fixed the problem that the property of the resource file and the property value in the bsradm dump output are different.
[BSR-617] Fixed an issue where the disk size was not updated when executing the bsrcon /info command after disk shrinking.
Improvement
[BSR-767] Fixed the problem of missing output of gi value in events2 of standby node when DR node is down
[BSR-766] Added P_ZEROES packet for processing REQ_OP_WRITE_ZEROES bio
[BSR-734] Add event output when split-brain occurs
1.6-A12
Release Date
Jun 3, 2021
Release notes
Bug
[BSR-763] Fix the problem that the repl. disk is not locked when rebooting after setting svc-auto-up no;
1.6-A11
Release Date
May 27, 2021
Release notes
Bug
[BSR-747] [Linux] Fixed inconsistency issue, when disk failure during replication or synchronization and recovery/initial synchronization is completed
Improvement
[BSR-756] Improved the problem of different levels of logs that are split and logged in windows events.
Task
[BSR-755] merged WDRBD 1.5.8.7 patch.
1.6-A10
Release Date
Apr 7, 2021
Release notes
Bug
[BSR-750] [Linux] Fix the problem of not responding when trying to mount on the primary node
[BSR-748] [Linux] Fix the problem that the bsr service cannot be run and reinstalled when the kernel module is updated in Ubuntu environment
[BSR-747] [Linux] Fixed inconsistency issue, when disk failure during replication or synchronization and recovery/initial synchronization is completed
[BSR-746] [Linux] Fix the problem of intermittent OS reboot when removing the primary disk during repl. and synchronization.
[BSR-744] [Linux] Fix oos remains when disk removal operation is happened during synchronous replication
[BSR-743] Fix the problem that the status of WFBitMapT/StartingSyncS does not change after invalidate command when operating multiple volumes
Improvement
[BSR-735] Improve --discard-my-data command
[BSR-734] Improved to output event when split-brain occurs