Enlarging a pool leads to 'No such file or directory'

[Christoph Haas]

While running a backup I realized that the pool had too few volumes (on disk). So I increased the number of volumes in the Pool definition, went to the bconsole and reloaded.

However that instantly lead to this error:

18-Dec 11:37 bareos-sd JobId 3605: Warning: stored/mount.cc:246 Open device "DiskVmwareStorage9" (/srv/stor2/bareos/vmware) Volume "VmwareIncr-2346" failed: ERR=stored/dev.cc:597 Could not open: /srv/stor2/bareos/vmware/VmwareIncr-2346, ERR=No such file or directory

There is indeed no such file. The most recent files are:

-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:21 VmwareIncr-2339
-rw-r----- 1 bareos bareos 10736369871 Dez 18 03:21 VmwareIncr-2331
-rw-r----- 1 bareos bareos 10736369879 Dez 18 03:21 VmwareIncr-2338
-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:22 VmwareIncr-2337
-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:23 VmwareIncr-2341
-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:23 VmwareIncr-2340
-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:24 VmwareIncr-2343
-rw-r----- 1 bareos bareos 10736369887 Dez 18 03:27 VmwareIncr-2342

My Pool definition:

Pool {
  Name = VmwareIncr
  Pool Type = Backup
  Recycle = yes                       # Bareos can automatically recycle Volumes
  AutoPrune = yes                     # Prune expired volumes
  Volume Retention = 10 days          # How long should the Incremental Backups be kept?
  Maximum Volume Bytes = 10G           # Limit Volume size to something reasonable
  Maximum Volumes = 400               # Limit number of Volumes in Pool
  Label Format = "VmwareIncr-"       # Volumes will be labeled "Incremental-<volume-id>"
  Recycle Oldest Volume = yes
}

The only way I can solve that is cancelling all active jobs (cancel all yes) and restarting them.

I'm on Bareos 23.1.1 on Debian 12. Has anyone else seen this? What am I doing wrong?

Cheers
 Christoph

[Sebastian Sura]

Hi Christoph

how did you "increase the number of volumes in the Pool definition" ? Did you just edit the pool configuration file and do a reload ?
Or did you use the `add` command to add new volumes directly ?

https://docs.bareos.org/TasksAndConcepts/BareosConsole.html#console-commands

>>> This command is used to add Volumes to an existing Pool. That is, it creates the Volume name in the catalog and inserts into the Pool in the catalog, but does not attempt to access the physical Volume. Once added, Bareos expects that Volume to exist and to be labeled.

Are you sure that the message was printed _after_ you added new volumes to the pool ? Maybe this message was printed first and you only added new volumes afterwards ?

Could you provide the full job log ?

Kind Regards
Sebastian Sura

Am 18.12.24 um 11:42 schrieb Christoph Haas:

--
You received this message because you are subscribed to the Google Groups "bareos-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bareos-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/bareos-users/aa0dcfd4-a51e-4019-b1a5-1f8b48dd137an%40googlegroups.com.

-- 
 Sebastian Sura                  sebasti...@bareos.com
 Bareos GmbH & Co. KG            Phone: +49 221 630693-0
 https://www.bareos.com
 Sitz der Gesellschaft: Köln | Amtsgericht Köln: HRA 29646
 Komplementär: Bareos Verwaltungs-GmbH
 Geschäftsführer: Stephan Dühr, Jörg Steffens, Philipp Storz

[Christoph Haas]

Thanks for the reply, Sebastian.

<tl;dr>Can't reproduce it.</tl;dr>

What I did (and what seemed to have worked this time):

1. Start the backup job to the "VmwareIncr" disk pool.
   
2. 18-Dec 14:21 bareos-dir JobId 3950: Start Backup JobId 3950, Job=vm-zabbix2.2024-12-18_14.21.39_12
   18-Dec 14:21 bareos-dir JobId 3950: Connected Storage daemon at bareos-dr.pmd5.org:9103, encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-dir JobId 3950:  Encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-dir JobId 3950: Probing client protocol... (result will be saved until config reload)
   18-Dec 14:21 bareos-dir JobId 3950: Connected Client: bareos-fd at localhost:9102, encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-dir JobId 3950:    Handshake: Immediate TLS
   18-Dec 14:21 bareos-dir JobId 3950:  Encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-dir JobId 3950: Pruning oldest volume "VmwareIncr-2110"
   18-Dec 14:21 bareos-dir JobId 3950: Pruning volume VmwareIncr-2110: 0 Jobs have expired and can be pruned.
   18-Dec 14:21 bareos-dir JobId 3950: Volume "VmwareIncr-2110" still contains jobs after pruning.
   18-Dec 14:21 bareos-dir JobId 3950: Using Device "DiskVmwareStorage1" to write.
   18-Dec 14:21 bareos-fd JobId 3950: Connected Storage daemon at bareos-dr.pmd5.org:9103, encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-fd JobId 3950:  Encryption: TLS_CHACHA20_POLY1305_SHA256 TLSv1.3
   18-Dec 14:21 bareos-fd JobId 3950: Extended attribute support is enabled
   18-Dec 14:21 bareos-fd JobId 3950: ACL support is enabled
   18-Dec 14:21 bareos-dir JobId 3950: Pruning oldest volume "VmwareIncr-2110"
   18-Dec 14:21 bareos-dir JobId 3950: Pruning volume VmwareIncr-2110: 0 Jobs have expired and can be pruned.
   18-Dec 14:21 bareos-dir JobId 3950: Volume "VmwareIncr-2110" still contains jobs after pruning.
   18-Dec 14:21 bareos-dir JobId 3950: Pruning oldest volume "VmwareIncr-2110"
   18-Dec 14:21 bareos-dir JobId 3950: Pruning volume VmwareIncr-2110: 0 Jobs have expired and can be pruned.
   18-Dec 14:21 bareos-dir JobId 3950: Volume "VmwareIncr-2110" still contains jobs after pruning.
   18-Dec 14:21 bareos-sd JobId 3950: Job vm-zabbix2.2024-12-18_14.21.39_12 is waiting. Cannot find any appendable volumes.
   Please use the "label" command to create a new Volume for:
       Storage:      "DiskVmwareStorage1" (/srv/stor2/bareos/vmware)
       Pool:         VmwareIncr
       Media type:   File
3. The pool:
   *list pool=VmwareIncr
   Using Catalog "MyCatalog"
   +--------+------------+---------+---------+----------+-------------+
   | poolid | name       | numvols | maxvols | pooltype | labelformat |
   +--------+------------+---------+---------+----------+-------------+
   |     34 | VmwareIncr |     203 |     203 | Backup   | VmwareIncr- |
   +--------+------------+---------+---------+----------+-------------+
   
4. Edit /etc/bareos/bareos-dir.d/pool/VmwareIncr.conf
5. Increase the number of "Maximum Volumes = …" to a higher number
   
6. bconsole
7. reload
8. 18-Dec 14:26 bareos-dir JobId 3950: Pruning oldest volume "VmwareIncr-2110"
   18-Dec 14:26 bareos-dir JobId 3950: Pruning volume VmwareIncr-2110: 0 Jobs have expired and can be pruned.
   18-Dec 14:26 bareos-dir JobId 3950: Volume "VmwareIncr-2110" still contains jobs after pruning.
   18-Dec 14:31 bareos-dir JobId 3950: Pruning oldest volume "VmwareIncr-2110"
   18-Dec 14:31 bareos-dir JobId 3950: Pruning volume VmwareIncr-2110: 0 Jobs have expired and can be pruned.
   18-Dec 14:31 bareos-dir JobId 3950: Volume "VmwareIncr-2110" still contains jobs after pruning.
9. The pool has picked up my change:
   *list pool=VmwareIncr
   +--------+------------+---------+---------+----------+-------------+
   | poolid | name       | numvols | maxvols | pooltype | labelformat |
   +--------+------------+---------+---------+----------+-------------+
   |     34 | VmwareIncr |     203 |     204 | Backup   | VmwareIncr- |
   +--------+------------+---------+---------+----------+-------------+
   
10. The job hasn't picked up the change yet:
    Running Jobs:
    Console connected at 18-Dec-24 14:32
     JobId Level   Name                       Status
    ======================================================================
      3950 Increme  vm-zabbix2.2024-12-18_14.21.39_12 is waiting for an appendable Volume
11. *mount
    The defined Storage resources are:
    1: DiskCritical
    2: DiskDatabase
    3: DiskVmware
    4: File
    5: Tape
    Select Storage resource (1-5): 3
    Connecting to Storage daemon DiskVmware at bareos-dr.pmd5.org:9103 ...
    3001 OK mount requested. Device="DiskVmwareStorage1" (/srv/stor2/bareos/vmware)
12. Suprisingly this worked. The job has picked up the enlarged pool and has taken the new volume:
    18-Dec 14:36 bareos-dir JobId 3950: Created new Volume "VmwareIncr-2351" in catalog.
    18-Dec 14:36 bareos-sd JobId 3950: Labeled new Volume "VmwareIncr-2351" on device "DiskVmwareStorage1" (/srv/stor2/bareos/vmware).
    18-Dec 14:36 bareos-sd JobId 3950: Moving to end of data on volume "VmwareIncr-2351"
    18-Dec 14:36 bareos-sd JobId 3950: Ready to append to end of Volume "VmwareIncr-2351" size=223
    

To be honest I am not sure what I did differently this time.

Is it generally a bad idea to do it like this? Would you suggest using "add" instead? Thanks!

…Christoph

[Sebastian Sura]

You should not use add to try and create new file volumes, as that command only adds them to the database, but does not actually create them.  If you want to create a file volume, just use the label command.  The joblog even asks you to use it: "Please use the "label" command to create a new Volume for:".  If you do that, the job should also automatically pick it up.

I could not reproduce your issue either, so I am not sure what happened there.

Kind Regards
Sebastian Sura

Am 18.12.24 um 14:55 schrieb Christoph Haas:

To view this discussion visit https://groups.google.com/d/msgid/bareos-users/63eadc52-7494-482d-b6d9-5f8d365217e5n%40googlegroups.com.