CHOROBA: jo, to sem moh napsat rovnou sry
bacula-hw - storage kde bezi proxmox, na tom:
vm bacula-dir - director baculy
vm bacula-sd - storage daemon, kde mam mountnuty raid
vm bacula-sd (s3) - storage daemon, kde je s3fs mount do aws jako offside backup, (+ v aws img os bacula-dir, jako ec2 instanci)
na directoru mam configy v jednom fajlu v /servers
@|"sh -c 'cat /etc/bacula/schedules/*.conf'"
@|"sh -c 'cat /etc/bacula/jobdefs/*.conf'"
@|"sh -c 'cat /etc/bacula/servers/*.conf'"
nejak takhle...
Client {
Name = cli-xxx
Address = xxx
FDPort = 9102
Catalog = bacula
Password = "xxx"
File Retention = 125 days
Job Retention = 6 months
AutoPrune = yes
}
FileSet {
Name = "fs-xxx"
Include {
Options {
signature = MD5
compression = GZIP
WildDir = "*/.logs"
Exclude = yes
}
File = /var/xxx
File = /opt/xxx
File = /opt/backup-bacula/xxx
}
}
Job {
Name = "job-xxx"
JobDefs = "JobFullAt2"
Client = "cli-xxx"
FileSet = "fs-xxx"
Pool = "pool-incremental-xxx"
Full Backup Pool = "pool-full-xxx"
Allow Mixed Priority = yes
Differential Backup Pool = "pool-differential-xxx"
Storage = "sto-xxx"
Client Run Before Job = "/usr/local/bin/pre_script.sh"
Client Run After Job = "/usr/local/bin/post_script.sh"
}
Job {
Name = "res-xxx"
Type = Restore
Client = "cli-xxx"
FileSet = "fs-xxx"
Storage = "sto-xxx"
Where = /tmp/bacula-restores
Messages = Standard
Pool = Default
}
Pool {
Name = "pool-full-xxx"
Pool Type = Backup
Recycle = yes
AutoPrune = yes
Volume Retention = 63 days
Use Volume Once = yes
Maximum Volumes = 5
LabelFormat = "FB-xxx-"
# Maximum Volume Bytes = 100G
}
Pool {
Name = "pool-incremental-xxx"
Pool Type = Backup
Recycle = yes
AutoPrune = yes
Volume Retention = 59 days
Use Volume Once = yes
Maximum Volumes = 60
LabelFormat = "IB-xxx-"
}
Pool {
Name = "pool-differential-xxx"
Pool Type = Backup
Recycle = yes
AutoPrune = yes
Volume Retention = 59 days
Maximum Volumes = 11
Use Volume Once = yes
LabelFormat = "DB-xxx-"
}
Storage {
Name = "xxx"
Address = xxx
SDPort = 9103
Device = "xxx"
Media Type = File
Maximum Concurrent Jobs = 3
Password = "xxx"
}
a vsechno co je nejak kriticka vec, tak ma na bacula-dir v /servers druhy konfig pro bacula-sd (s3), takze se pro to spousti fakticky dva joby v rozdilnych intervalech, s rozdilnou retenci, kazdy na jinou storage...