Hello,
Running 2 jobs with same priority at the same is running well.
So I used the priority to order the start of the jobs and try to run
different priority jobs by using "Allowed Mixed Priority = yes" for
all the relevant jobs:
# Job definition for rotate
JobDefs {
Name = "RotateJob"
Type = Backup
Level = Incremental
Schedule = "Cycle"
Max Start Delay = 1240000
Max Wait Time = 3600000
Max Run Time = 3600000
Spool Data = yes
Spool Attributes = yes
Messages = Standard
Pool = Daily
Storage = "Disk2"
Incremental Backup Pool = Daily
Differential Backup Pool = Weekly
Full Backup Pool = Monthly
Rerun Failed Levels = yes
Allow Duplicate Jobs = no
Cancel lower level duplicates = yes
Cancel Queued Duplicates = yes
Cancel Running Duplicates = no
Accurate = yes
Priority = 9
Allow Mixed Priority = yes
}
…
Here for each priority one job def:
…
Job {
Name = "conny_full"
JobDefs = "RotateJob"
Client = conny-fd
FileSet = "Full Set"
Write Bootstrap = "/var/lib/bacula/conny_full.bsr"
}
Job {
Name = "conny_home"
JobDefs = "RotateJob"
Client = conny-fd
FileSet = "home Set"
Write Bootstrap = "/var/lib/bacula/conny_home.bsr"
Priority = 13
}
Job {
Name = "nihilnihil_home"
JobDefs = "RotateJob"
Client = nihilnihil-fd
FileSet = "home wo WinWork Set"
Write Bootstrap = "/var/lib/bacula/nihilnihil_home.bsr"
Priority = 12
}
…
By the way only nihilnihil_home is running for the moment although:
*status dir
backup-dir Version: 9.4.2 (04 February 2019) x86_64-pc-linux-gnu debian
buster/sid
Daemon started 21-May-20 22:27, conf reloaded 21-May-2020 22:27:40
Jobs: run=560, running=6 mode=0,0
Heap: heap=913,408 smbytes=378,789 max_bytes=951,271 bufs=1,227 max_bufs=1,521
Res: njobs=40 nclients=17 nstores=6 npools=10 ncats=1 nfsets=8 nscheds=3
Scheduled Jobs:
Level Type Pri Scheduled Job Name Volume
===================================================================================
Incremental Backup 9 06-Jul-20 23:50 backup_full DISK005
Incremental Backup 9 06-Jul-20 23:50 nihilnihil_full DISK005
Incremental Backup 9 06-Jul-20 23:50 conny_full DISK005
Incremental Backup 9 06-Jul-20 23:50 mail_full DISK005
Incremental Backup 9 06-Jul-20 23:50 file_full DISK005
Incremental Backup 9 06-Jul-20 23:50 web_full DISK005
Incremental Backup 12 06-Jul-20 23:50 nihilnihil_home DISK005
Incremental Backup 13 06-Jul-20 23:50 web_server DISK005
Incremental Backup 13 06-Jul-20 23:50 file_home DISK005
Incremental Backup 13 06-Jul-20 23:50 conny_home DISK005
Incremental Backup 13 06-Jul-20 23:50 mail_home DISK005
Full Backup 19 06-Jul-20 23:51 BackupCatalog DISK002
====
Running Jobs:
Console connected using TLS at 26-Jun-20 06:37
JobId Type Level Files Bytes Name Status
======================================================================
48916 Back Full 209,045 61.50 G nihilnihil_home is running
48917 Back Full 0 0 web_server is waiting for higher
priority jobs to finish
48918 Back Full 0 0 file_home is waiting execution
48919 Back Full 0 0 mail_home is waiting execution
48920 Back Full 0 0 conny_home is waiting execution
48921 Back Full 0 0 BackupCatalog is waiting execution
====
The limit of max 2 jobs is setup in the drives configuration area.
…
Director { # define myself
Name = backup-dir
Description = "Director on backup server backup."
DIRport = 9101
QueryFile = "/etc/bacula/scripts/query.sql"
WorkingDirectory = "/var/lib/bacula"
PidDirectory = "/run/bacula"
Maximum Concurrent Jobs = 4
Password = "password"
Messages = Daemon
DirAddress = backup
# TLS configuration
TLS Enable = yes
TLS Require = yes
TLS Verify Peer = yes
TLS Allowed CN = "bacula@backup"
TLS Allowed CN = "bacula@in94"
TLS Allowed CN = "bacula@nihilnihil"
TLS Allowed CN = "pierrei@"
@/etc/bacula/tls_server.conf
}
…
Storage { # definition of myself
Name = backup-sd
SDPort = 9103 # Director's port
WorkingDirectory = "/var/lib/bacula"
Pid Directory = "/var/run/bacula"
Maximum Concurrent Jobs = 20
SDAddress = backup
# Incoming connections from director
TLS Enable = yes
TLS Require = yes
# No Verify because of sd connection cookie
TLS Verify Peer = no
# Server port
@/etc/bacula/tls_server.conf
}
…
Storage {
Name = "FibreCAT TX48 S2"
address = backup
SDPort = 9103
Password = "password"
Device = "FibreCAT TX48 S2"
Media Type = "LTO-3"
Autochanger = Yes
Maximum Concurrent Jobs = 2
# TLS Configuration
TLS Enable = yes
TLS Require = yes
@/etc/bacula/tls_client.conf
}
…
So I would be expected that if only one job of prio 9 is running the job
nihilnihil_home
would be started. If the last prio 9 job ended the next prio 12 job would be
started.
By the way, as shown above the job web_server is still waiting and the als shown
below from the logs the running job nihilnihil_home wasn't already started
before all prio 9 has been ended.
The prio 9 jobs has been running in parallel with max of 2 concurent jobs:
Job: backup_full.2020-07-05_23.50.00_21
Scheduled time: 05-Jul-2020 23:50:00
Start time: 05-Jul-2020 23:54:15
End time: 06-Jul-2020 01:30:36
Elapsed time: 1 hour 36 mins 21 secs
Priority: 9
Job: web_full.2020-07-05_23.50.00_22
Scheduled time: 05-Jul-2020 23:50:00
Start time: 05-Jul-2020 23:54:15
End time: 06-Jul-2020 00:49:51
Elapsed time: 55 mins 36 secs
Priority: 9
Job: file_full.2020-07-05_23.50.00_23
Scheduled time: 05-Jul-2020 23:50:00
Start time: 06-Jul-2020 00:49:58
End time: 06-Jul-2020 01:30:36
Elapsed time: 40 mins 38 secs
Priority: 9
Job: conny_full.2020-07-05_23.50.00_25
Scheduled time: 05-Jul-2020 23:50:00
Start time: 06-Jul-2020 01:30:47
End time: 06-Jul-2020 03:52:43
Elapsed time: 2 hours 21 mins 56 secs
Priority: 9
Job: mail_full.2020-07-05_23.50.00_24
Scheduled time: 05-Jul-2020 23:50:00
Start time: 06-Jul-2020 01:30:44
End time: 06-Jul-2020 02:07:52
Elapsed time: 37 mins 8 secs
Priority: 9
You can see at the time backup_full was running in meantime the two jobs
web_full and file_full has been running.
I wonder me that file_full has same endtime with backup_full, but maybe this
is a limit of storing the attrs to the database? This is not really relevant.
conny_full and mail_full has been started and running in parallel so
I would expect that nihilhihil_home should be started after 02:07:52,
but it has been really started after 03:52:43 so it was waiting for the
end of the last prio 9 job mail_full.
Any idea why the allow mixed priority = yes rule is not working well?
I'm unsure but I mean with older release 7.x before upgraded to buster
it has been worked well.
Cheers,
Pierre
_______________________________________________
Bacula-users mailing list
Bacula-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bacula-users
