Bacula-users

[Bacula-users] Parallel copy jobs revisited

2012-06-04 08:55:02
Subject: [Bacula-users] Parallel copy jobs revisited
From: Uwe Schuerkamp <uwe.schuerkamp AT nionex DOT net>
To: Bacula Users Mailing List <bacula-users AT lists.sourceforge DOT net>
Date: Mon, 4 Jun 2012 14:53:02 +0200
Hi folks,

this is my 2nd attempt to configure bacula 5.2.6 to run more than one
copy job at a time, and for the life of me I cannot find the error in
my config. Parallel backup jobs (both full and incr.) work like a
charm, but for some reason bacula refuses to run more than one copy
job from disk to tape... I'm getting desparate to solve this one
because the long queue of copy jobs just gets, well, longer ;-) 

Jobs are copied from the disk-based "FileStorage_Full" storage to the
"offline" pool (LTO5). 

Here's my current config: 

copy job definition: 

Job {
  Name = "copy"
  Type = Copy
  Level = Full
  Priority = 15  # make sure all other jobs are finished first                 
  Pool = Online_full
  Client = deniol186-fd
  FileSet = "Full Set deniol186"
  Messages = Standard
  Maximum Concurrent Jobs = 8 # changing this number has no effect
  Schedule = DailyCopy
  Selection Type = PoolUncopiedJobs
}

Storage "lto4": 

Storage {
  Name = lto4
  Address = tapeserver
  SDPort = 9103
  Password = "XXXXXX"
  Device = lto4
  Media Type = LTO-4
  Autochanger = yes
  AllowCompression = No
  Maximum Concurrent Jobs = 8
}

FileStorage_Full: 

Storage {
  Name = FileStorage_full
  Address = tapeserver
  here      
  SDPort = 9103
  Password = "xxxxxx"
  Device = FileStorage_full
  Media Type = File
  Maximum Concurrent Jobs = 8
}

In the client definition, I have 

Client {
  Name = tapeserver
  Address = tapeserver
  FDPort = 9102
  Catalog = MyCatalog
  Maximum Concurrent Jobs = 8
  Password = "XXXXXX"
  File Retention = 30 days            # 30 days                                
  Job Retention = 6 months            # six months                             
  AutoPrune = yes                     # Prune expired Jobs/Files               
}

My director allows up to 20 concurrent jobs, same goes for the FD on
the backup server: 

Director {                            # define myself                          
  Name = tapeserver-dir
  DIRport = 9101                # where we listen for UA connections           
  QueryFile = "/server/bacula/etc/query.sql"
  WorkingDirectory = "/server/bacula/var/bacula/working"
  PidDirectory = "/var/run"
  Maximum Concurrent Jobs = 20
  Password = "XXXXX" 
  Messages = Daemon
}

What am I missing? Is it impossible to run more than one copy job at
once? I seem to be limited by the number of readers / writers on the
disk volume it seems: 

Jobs waiting to reserve a drive:
   3602 JobId=27199 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27201 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27205 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27207 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27209 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27211 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).
   3602 JobId=27213 device "FileStorage_full" (/mnt/msa/online_backup)
   is busy (already reading/writing).


Used Volume status:
OFFLINE07_02 on device "lto4" (/dev/nst0)
    Reader=0 writers=1 devres=0 volinuse=1
full-0074 on device "FileStorage_full" (/mnt/raid/online_backup)
    Reader=1 writers=0 devres=0 volinuse=1
full-0074 read volume JobId=27199
full-0074 read volume JobId=27201
full-0074 read volume JobId=27205
full-0074 read volume JobId=27207
full-0074 read volume JobId=27209
full-0074 read volume JobId=27211
full-0074 read volume JobId=27213
full-0074 read volume JobId=27221

However it should be possible for more than one *reader* to access the
disk based volumes? I'm really confused why this setup won't work when
parallel backup jobs write to a single volume just fine. 

I've also tried increasing the max. concurrent jobs in the "Device"
definitions, sadly to no avail. 

All the best & thanks in advance for your help, 

Uwe 

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Bacula-users mailing list
Bacula-users AT lists.sourceforge DOT net
https://lists.sourceforge.net/lists/listinfo/bacula-users

<Prev in Thread] Current Thread [Next in Thread>