Release Notes for Bacula Enterprise 10.2

The Director (Dir) and all Storage Daemons (SDs) must all be upgraded to version 10.2 because they are not compatible with prior versions.

Catalog format in version 10.0 and greater:

The previous major release 10.0.0 of Bacula uses a new catalog format. We provide a set of scripts that convert 6.6.x and earlier catalog versions to the new 10.0.0 format (1020). Please see the white paper “Migration-to-10.2” for detailed instructions on the migration process.

The database upgrade from 1019 to 1020 is fast and simple for most users.

The Bacula Enterprise RPM spec files have been modified to detect the database change and do the upgrade for you (from 8.2, 8.6 and 8.8). We strongly recommend that you save a copy of your existing database before upgrading using RPMs. If you want to have more control over the catalog upgrade step when upgrading using RPMs, you may want to contact the Bacula Systems support team to get the update_bacula_tables script and run it before the RPM installation step.

If you upgrade BWeb Management Suite from 8.x, you must update the BWeb SQL tables with the script upgrade-6.0-8.8_postgresql.sql or upgrade-6.0-8.8_mysql.sql.

If you are starting from scratch, you don’t need to run the update_bacula_tables script because the create_bacula_tables script automatically creates the tables using the new format.

Restricted Console users must modify their existing Console resources to specify DirectoryACL=*all* and UserIdACL=*all* in order to continue to use the restore command. If these two directives are not set, the restore command will return an error.

On Suse Linux, the Director and the Storage Daemons are now started with the unix account “bacula”. To upgrade a Suse system to Bacula Enterprise 8.8 or later, the ownership of the Bacula volumes and the configuration files must be changed to “bacula:bacula” and the PostgreSQL/MySQL configuration may have to be adapted. More information can be found in the Migration-to-10.2 whitepaper available in your download area.

Debian Sarge/Etch/Lenny/Squeeze, Ubuntu Hardy/Lucid, Redhat 5, SLES 10, SLES 11, SLES 11.1, SLES 11.2 are no longer supported in version 10.2. Old filedaemons are still compatible and can be used to protect these systems.

If you are upgrading from 8.x, please read the ReleaseNotes from the last 10.0 version.

Release 10.2.4 17 May 2019

10.2.4 will be a minor bug fix release. There are a few new features.

  • Fix #5041 about the Copy of an Incomplete job failing with the error

  • XXX digest not same FileIndex=YYYY

  • Fix #5053 about BVFS commands not compatible with Console ACLs wildcards

  • Fix #4754 about incorrect path for the psql command in grant_postgresql_privileges

  • Fix #4778 about incorrect comment in delete_catalog_backup script

  • Fix #4800 - Fix OS X Readme

  • Fix 4599 bools.py reports VacuumBadRef RefTooSmall

  • Fix Azure compilation on Debian Stretch

  • Skip empty lines when generating the FileSet from a command

  • Update ./configure to search librsync in lib64

  • Add filename and digest to message “SHA1 digest not same FileIndex=XXX”

  • Add block address to bls -v output

  • Add AllowIncompleteJobs Job directive

  • linuxbmr: update Bacula-rescue.sh to version 2.2.1

  • bweb: Allow setting accurate directive as –directive parameter in scan_datacenter.pl script

  • bweb: Fix #4594 add missing headers sending in cgi scripts

  • bweb: Fix #4696 add WormCommand directive support to device configuration

  • bweb: Fix #4744 javascript security error when ad-block type plugin is used

  • bweb: Fix #4747 system auth option to not set setuid before bconsole commands

  • bweb: Fix #4748 BWeb was showing delete client button for users without r_configure role assigned

  • bweb: Fix #4757 empty page on run job wizard

  • bweb: Fix #4771 add missing ACL roles for MySQL and SQLite catalog databases

  • bweb: Fix #4773 update copyright and license statements

  • bweb: Fix #4791 language setting not working for particular users

  • bweb: Fix #4796 problem with creating new user if sqlite catalog database is used

  • bweb: Fix #4804 missing client pie charts if sqlite catalog database is used

  • bweb: Fix #4817 re-compute media button in restore wizard if selected path doesn’t belong to fileset path

  • bweb: Fix #4830 show default volume retention value on pool list page if no volume retention defined in config

  • bweb: Fix #4943 failed jobs link on dashboard to show all job types

  • bweb: Fix #4954 remove extra space in Last Jobs header bar

  • bweb: Fix #4968 allow plus character in path and password validators

  • bweb: Fix error after successful system auth login when no page action is set

  • bweb: Request #4188 add description field to information box header on main page

  • bweb: Request #4775 add XenServer Hypervisor plugin support

  • bweb: Update PT translation. Thanks to Heitor

  • cloud: Add check for empty bucket name

  • cloud: Fix some fprintf() formats

  • cloud: Standalone WAS driver implemented

  • cloud: Update azure version

  • cloud: Add stream I/O to libwas API

  • cloud: Allow use of file path for WAS driver

  • cloud: Compute part hash at upload and download

  • cloud: Fix protocol and uri style detection

  • cloud: Generic drive sets azure params

  • cloud: Update was_driver ouput and error code

  • docker: Fix #4956 about a segfault issue when Docker is not available

  • mssql: Fix #2387 Add directives for Recovery Model

  • mysql: Display an error if the BINLOG list command has an error

  • ndmp: Add C version of idx_dump.pl

  • ndmp: Display an error if the history file analysis has a problem

  • ndmp: Fix #3871 Add support for UnityOE DELL/EMC systems

  • ndmp: Fix #4972 About a bug in the history file restore handling

  • ndmp: Fix issue with NDMP config backup

  • ndmp: Improve volume detection and fix error message if the NDMP system is not available

  • ndmp: Check if custom_ndmp_idx_dump program exists to analyze the history file

  • ndmp: Mitigate the restore status with the NDMP log messages

  • rhv: FEAT: Detect incompatibility with vms derived from dependent templates with CLONE BACKUP

  • rhv: FEAT: Compatibility with iSCSI Storage Domain

  • rhv: FIX: Some bugs when creating a new entity

  • rhv: FIX: Wait time in some operations and check for cancel in Bacula

  • rhv: STYLE: Message when any target is found.

  • rhv: FEAT: Notification in case snapshot removal failed

  • rhv: FIX: Minor bugs in JSON result

  • rhv: STYLE: Change comments in several logs

  • rhv: FEAT - Backup and restore of templates

  • rhv: STYLE - Change plugin version to 3.0.0

  • rhv: FEAT - Add template option in Bacula restore

  • rhv: FEAT - Extend time to remove snapshot to 15 min

  • rhv: FEAT - Add warning when a RHV bug happens while removing snapshot

  • rhv: Sync with 10.0 multiple merge operations

  • rhv: 2.0.1 Add patch number in plugin version and fix option ‘insecure’

  • rhv: Add support for Storage Domain with type storage SAN and add logs about upload/download/speed

  • rhv: Implement backup templates

  • rhv: Improve sleep times for insecurity of SDK

  • rhv: Added sizes in config file json

  • rhv: FIX: Minor bugs in messages

  • rhv: FEAT: Add list version and check compatibility of clusters and storage to restore VM in a different cluster/storage.

  • rhv: FEAT: Optimize data retrieval for the server in the “list” operation.

  • rhv: FIX: Get description and ID of snapshots in the “list” operation

  • rhv: [FIX] Default action in list Bacula

  • rhv: [FEAT] Try N times before throwing an error during the first connection with the RHV server

  • rhv: [FEAT] In “list” command, the result shows the type of element that is listed

  • rhv: [FEAT] Improve the list format when VMs are shown

  • rhv: [FEAT] Detect compatibility of each VM backup and determine which backup operation (Backup clone or Backup snapshot) will be applied to each VM.

  • rhv: [FEAT] When the backup type is changed, the plugin will log a warning about the situation

  • rhv: [FIX] Minor bugs in restore VM when the VM will be restored in a different storage

  • rhv: [DOC] Add table to explain behaviour of plugin depending on the hardware configuration

  • rhv: [DOC] Fix some bugs in documentation and added new features with Truth Table

  • rpms: Adapt SUSE spec files for new was driver

  • rpms: Adapt cloud spec file for new was driver

  • rpms: Add Sybase spec for SUSE

  • rpms: Fix #4767 - Fix systemd for BWeb on SLES

  • rpms: Fix #4823 - Fix Docker package dependency

  • rpms: Remove librsync-devel package requirement on SUSE platforms

  • rpms: Use new ndmp_idx_dump program instead of idx_dump.pl

  • rpms: Fix BWeb spec for SLES

  • vsphere: Add a retry loop around the PrepareForBackup() call

  • vsphere: Do not retry the call to PrepareForAccess() if the flag is already set in vCenter/VM

  • vsphere: Fix #5019 about snapshot creation error with long Job names

Bugs fixed/closed since last release: 2387 3871 4188 4594 4642 4696 4744 4747 4748 4754 4757 4767 4771 4773 4775 4778 4791 4796 4800 4804 4817 4823 4830 4943 4954 4956 4968 4972 5019

Release 10.2.3 05 March 2019

10.2.3 is a minor bug fix release.

  • Allow running a Virtual Full on a single Job

  • Display daemon pid in .apiV2 status output

  • Scan volume byte per byte after a record error to find the next valid record with the -p option

  • Abort a job if the attribute/data/checksum sequence is incorrect

  • Add Jmsg after a verify error

  • Add device device_specific_close() function to fix open_device() code

  • Add messages for Incomplete Jobs

  • Change endblock edit to unsigned – suggested by Martin Simmons

  • Check JobMedia validity after an incomplete job

  • Fix bscan change volume status from Purged/Recycled to Archive if we recover a Job

  • Fix #3225 about Migration issue when the Job resource is no longer defined

  • Fix #4598 Display JobIds used in the restore job log

  • Fix #4621 Use more appropriate computation for VolIndex when creating restore .bsr

  • Fix #4675: bscan segfault on DEDUP volumes

  • Fix #4676 about deadlock when reloading the configuration after a SIGHUP

  • Fix Daemon message “Message repeated X times” count

  • Fix bscan -p parameter in getopt()

  • Fix creation of bad JobMedia records in Incomplete Job

  • Fix file:block display on some error messages

  • Fix misplaced cancel check reported by Alain

  • Fix xattr on CIFS share

  • Fix minor memory leak in error condition when reading the spool file

  • aligned: Fix #4547 about orphan file descriptors after the truncate command

  • bweb: Add default table sorting option to bsTable options

  • bweb: Add filter option to restore wizard to include jobs in error

  • bweb: Change Sybase SBT plugin name into Sybase ASE SBT plugin name

  • bweb: Fix #4545 calendar icon too large and button label in VMware restore wizard

  • bweb: Fix #4548 VMware restore when raw disks are used on VMware server

  • bweb: Fix #4561 Swift plugin option abort_on_error not working correctly

  • bweb: Fix #4567 added option to remove admin job in can_datacenter.pl vmware integration tool

  • bweb: Fix #4570 VMware plugin icon too large

  • bweb: Fix #4581 display correct media expiration date and time

  • bweb: Fix #4590 incorrect job log order

  • bweb: Fix #4592 clearing worksets with scan_datacenter.pl when commit_and_reload option is used

  • bweb: Fix #4597 add warning on job rename when SAP/OracleRMAN plugin is defined in fileset

  • bweb: Fix #4602 pie chart labels overlap and are not readable

  • bweb: Fix #4620 using filters with client group pie charts

  • bweb: Fix #4635 remove visible resource pools selection for ESXi host in VMware resource browser

  • bweb: Fix #4660 capability to run any job by user with assigned restricted client group

  • bweb: Fix #4663 about adding a comment on lighttpd SSL configuration in httpd.conf

  • bweb: Fix #4705 extend restore wizard capability to restore jobs terminated with error

  • bweb: Fix #4711 wrong labels in client OS type pie chart

  • bweb: Fix getting VMware disks for non-flat file formats

  • bweb: Hide number of items field on group statistics page

  • bweb: Request #4533 default sort job list by time (descending)

  • bweb: Request #4533 move the apply job filters button at the top of box

  • bweb: Request #4710 add a message about required Samba common tools to push config script

  • bweb: Update job status icons

  • bweb: Fix #4721 about cancel and stop buttons not using selected running jobs

  • cloud: Fix volume list output in BCloud

  • cloud: Get information about the cloud driver with high debug level

  • mssql: Add a “driver” plugin command keyword to overwrite the ODBC driver name

  • mssql: Add plugin command “driver” keyword support

  • mssql: Fix #4454 about an empty selection warning with system db

  • mssql: Fix #4455 about incorrect replica message for sysdb

  • mssql: Fix #4522 Avoid infinite wait in GetConfiguration

  • mssql: Fix #4522 loosen the timeout to 1 minute

  • mssql: Fix #4523 About MSSQL “copyonly” value that is not taken into account correctly

  • ndmp: Add Single Item Restore support for filenames that contain spaces

  • postgres: Fix issue with createdb.sql generation

  • registration wizard: Add missing line to define OPENVPN_LOCATION on Windows systems

  • registration wizard: Fix #4497 do not setup ovpn when usevpn is false. Add version to app.

  • rest-api: Fix #4497 registering client without VPN option set

  • dedup: Display a list of problematic jobs after the list of files with btools.py

  • rhv: Added Content-Length GET actualSize or ProviosionedSize

  • rhv: Add libraries and sdk 4.3

  • rhv: Add download performance log messages

  • rhv: Download and restore using the Content-Length

  • rhv: Fix #4642 about ‘Accurate mode is not supported’ message

  • rhv: Handle error 404. Not working as expected

  • rpms: Fix #4519 about libexpat dependency issue in vSphere-vix package 6.5

  • rpms: Fix #4714 about incorrect URL field

  • rpms: Fix mt-st dependency for OpenSuse Leap 15

  • rpms: Fix OpenSSL dependencies for SLES and OpenSuse Leap

  • rpms: Fix rpm spec files for SLES 15 and OpenSuse Leap 15

  • rpms: Fix tokyocabinet build for SLES and OpenSuse

  • rpms: Fix vSphere spec for Suse

  • rpms: Fix #4519 on Suse Linux and OpenSuse

  • vsphere: Fix #3211 Add force_san plugin command line option

  • vsphere: Fix #4639 Display the selected transport mode in Job output

  • win32: Fix #4644 VSS Plugin will report an error if the System Writer is not available for SYSTEMSTATE

  • win32: Update openssl library to 10.0.2r

  • winbmr: Quiet some minor errors about disk detection

  • linuxbmr: Update plugin to 2.2.1 for UEFI and SLES

Bugs fixed/closed since last release: 3211 3225 4454 4455 4497 4519 4522 4523 4533 4545 4547 4548 4561 4567 4570 4581 4590 4592 4597 4598 4602 4620 4621 4635 4639 4642 4644 4660 4663 4675 4676 4705 4710 4711 4714 4519 4721

Release 10.2.2 09 January 2019

10.2.2 is a minor bug fix release.

  • Add procedure to upgrade the catalog from version 16 to 1020

  • Add bconsole update command to set the Job comment field

  • Fix #4433 about ‘UPDATE File SET MD5=’…’ WHERE FileId=0’ error when using SpoolAttributes=no

  • Fix #4449 about an incorrect pool selected with the restart command

  • Fix bug #4379 certain fields of Media record not reset after Truncate command

  • Fix memory issue with the strip fileset option

  • Improve volume truncation error messages

  • bweb: Add support for Sybase SBT plugin

  • bweb: Fix #4518 Azure endpoints validation

  • bweb: Fix #4526 copy function on storage autochanger config page not working

  • bweb: Add a note to ‘backup multiple virtual machines wizard’ about required unique template names

  • bweb: Fix #4451 add support for the same device defined in different storage resources

  • bweb: Fix #4502 20th day of month setting in schedule resource

  • bweb: Request #4493 add job status description to the running job box

  • bweb: Fix #4529 add validation for server name field in “VMware center”

  • cloud: Fix #4457 about a harmless message “Catalog VolCloudParts mismatch” with the bconsole ‘cloud list’ command

  • mssql: Display connection string in JobMessage

  • mssql: Do not display the password in debug trace file

  • mssql: Fix #4455 incorrect replica message for sysdb

  • mssql: Fix #4454 don’t display empty selection warning with system db

  • ndmp: Fix #4524 Add plugin timeout option. Default is 300s

  • ndmp: Fix #4524 about segfault during the cleanup after the backup

  • ndmp: Include more NDMP messages in the job log, including connection errors

  • rhv: Fix #4499 about the issue “Cannot transfer Virtual Disk: The following disks are locked”

  • rhv: Fix #4487 about protocol error message

  • vsphere: Update VIX library to 6.5

  • win32: Upgrade to OpenSSL v1.0.2q

  • rpms: Fix #4513 - Missing scan_datacenter.pl from SLES BWeb package

  • rpms: Add new spec file for Sybase plugin

  • rpms: Fix #4360 - storagectl.conf overwritten by upgrade

  • rpms: Fix vsphere package with VIX 6.5

  • rpms: Upgrade VMware VIX version

  • debs: Fix #4439 about bsysreport.pl script missing

  • docs: Remove references to Python scripting

  • freebsd: workaround for unimplemented MAP_NORESERVE in FreeBSD

  • kvm: Fix #4475 about a restore issue with the message “Zero length filename”

  • kvm: Fix root directory backup name

  • rest-api: Change using client address into using client name in vpn certificate

  • xenserver: Fix unexpected truncation of an error message while deleting a snapshot

Bugs fixed/closed since last release: 4360 4379 4433 4439 4449 4451 4457 4475 4493 4502 4513 4518 4524 4526 4499 4487 4455 4454 4529

Release 10.2.1 30 November 2018

10.2.1 is a minor bug fix release. There are also a few new features.

  • Add ‘prune jobs/files all’ command

  • bweb: Add support for multiple resource pools or datastores selection in vmware resource browser

  • bweb: Request #4402 Add full text log search field

  • proxmox: Fix #4369 Add support for pve-manager/5.2-1

  • Fix #4386 About incorrect permission on directories after a restore with replace=ifnewer

  • Fix MaxVolumeBytes accounting after a mount request

  • Fix verify volume jobs with sparse files

  • bcloud: Make chown OPENVPN_USER:GROUP optional on client files

  • registration wizard: Fix #4372 About incorrect OpenVPN detection

  • bweb: Fix #4350 showing validation error message at the top of the page

  • bweb: Fix #4363 negative duration being displayed for queued jobs

  • bweb: Fix #4366 add –remove_jobs option description to scan_datacenter.pl help message

  • bweb: Fix #4398 filters in scan_datacenter.pl script for multiple virtual machines

  • bweb: Fix #4423 remove CEPH driver setting from cloud add/edit page

  • bweb: Request #4086 add to installation instruction commands to check if the vSphere Automation Perl SDK is installed correctly

  • bweb: Request #4333 make menu and button labels consistent

  • bweb: Request #4350 move description text to page title

  • docs: Document new “prune all” feature

  • docs: Fix #4365 about two missing options in the bls man page

  • mysql: Fix incremental binary backup command

  • mysql: Fix segfault during backup

  • rest-api: Change default OpenVPN client config directory to /etc/openvpn/ccd

  • rest-api: Change default user:group to openvpn:openvpn in VPN post config script

  • rest-api: Fix #4370 default values for sysconfig directory

  • rest-api: Fix #4391 add netmask to ifconfig-push option (fix made by Aleksandar)

  • rest-api: Fix #4399 missing bottom part of bcloud page after saving job

  • rest-api: Fix setting privileges in VPN post config script

  • win32: Fix #4286 about the restore of a volume root directory with the hidden & system attributes

  • win32: Fix #4349 about a crash during restore in the heartbeat thread

Bugs fixed/closed since last release: 4086 4286 4333 4349 4350 4363 4365 4366 4369 4370 4372 4386 4391 4398 4399 4423 4402

Release 10.2.0 13 November 2018

10.2.0 is a major release.

New Features:

  • Bacula Daemon Real-Time Statistics Monitoring

  • Oracle S3 Cloud driver support

  • Google Cloud driver support

  • ACSLS Tape changer support

  • Add support for Windows security attributes backup over CIFS

Plugin:

  • Redhat Virtualization System Plugin (rhv)

  • MySQL Percona Plugin enhancements (prepare=sd/fd options)

  • LDAP Plugin AD Tombstone recovery support

BWeb Features:

  • New BWeb Management Suite look & feel

  • New Autocommit option

  • Elements in each page are now loaded asynchronously

  • Errors and warnings are now colorized in the job log window

  • The “run job” wizard has been improved

  • VMware tags support in the scan_datacenter tool

  • Add a logout button when using HTTP basic authentication

  • Add download “bsys_report” option

Misc Features:

  • Add Console ACLs glob pattern support

  • Add support for Windows files with non UTF16 names

  • Add VPN support for BCloud Service

  • Add Index support for Aligned Volumes (vSphere and Exchange plugins)

  • Add support for WORM tapes

  • Add support for SLES15, OpenBSD, Slackware 13

Important Changes:

  • Improvements to Snapshot management on Linux and Windows

  • Swift backup performance improvements

  • Add M_SECURITY messages when incoming connection is bad

  • Add Windows snapshot support to the delta plugin

More information can be found in the “New Features” section of the main manual.

Known Issues and Workarounds:

#1723 The hardlinks=yes FileSet option is not compatible with VirtualFull jobs.

This incompatibility might lead to problems during a restore. The error message displayed during a restore job is:

Error: create_file.c:327 Could not hard link /path: ERR=No such file or directory

To restore the file correctly, it is required to select the first occurrence of the hardlink file in the backup manually. The procedure to distinguish hardlinks in the catalog is available from Bacula Systems support team.

#2367 Virtual Full based only on Incremental/Differential issue with Accurate mode.

VirtualFull jobs based only on Incremental/Differential jobs do not carry forward information about deleted files in Accurate mode. The restore of a Full job + the VirtualFull job will restore files that were deleted after the Full job.

#4957 VirtualFull, Copy/Migration jobs based on Incomplete jobs - potential issue.

Under some circumstances a Copy/Migration/VirtualFull job may not be able to copy the data from a Job that was interrupted and restarted using the Incomplete Job feature. In version 10.2.3 and above, the Copy/Migration/VirtualFull job is aborted with the following error message:

Fatal error: catreq.c:691 SHA1 digest not same FileIndex=4 as attributes FI=3 or Fatal error: catreq.c:691 MD5 digest not same FileIndex=4 as attributes FI=3

If such an error message is logged, version 10.2.3 and above fails the job encountering this issue. The fix for this issue is to re-execute the Job that was interrupted to copy/migrate the data and not restart the incomplete job.

Verify jobs are strongly recommended with the Progressive Virtual Full technology.

Any customer having the error above should contact the Support Team to fully qualify the situation.