search

zmanda / amanda

Amanda Network Backup
https://www.zmanda.com/downloads/
Other
214 stars 107 forks source link

Only the first backup in disklist is made #253

Open ndias opened 5 months ago

ndias commented 5 months ago

Because the issue https://github.com/zmanda/amanda/issues/222 was closed and I can«t open again, and the suggest is to open a new issue.

The issue at @https://github.com/zmanda/amanda/issues/222 is a bug that should be resolved, it's not true this

"Due to multiple DLE entries, amanda is using degraded mode for the backups. Using a holdingdisk is mandatory in degraded mode."

Like I wrote in the issue, if i use multiple DLEs and a physical TAPE the holdding disk is not necessary and amanda doesn't complain about not having an holding disk.

From the man pages ... "The amanda.conf file may define one or more holding disks used as buffers to hold backup images before they are written to tape."

may ... not MUST so it's optional.

konidev20 commented 5 months ago

Hey @ndias,

We will check this out and see if there are any gaps in the documentation between 3.5.3 and 3.5.1. Thank you for your patience.

tanstafel commented 2 weeks ago

Just stumbled of the same issue. We upgraded from a CentOS 7 with amanda 3.3.3 to a Rocky9 with amanda 3.5.3 and our backup shows the same issue. We actually have a small holding disks to buffer some of the data from remote servers but most of the data to backup is already local. Now making the holdingdisk as large as the whole backup is ridiculous. This can't be right. Please provide a solution to this issue.

tanstafel commented 2 weeks ago

Just stumbled of the same issue. We upgraded from a CentOS 7 with amanda 3.3.3 to a Rocky9 with amanda 3.5.3 and our backup shows the same issue. We actually have a small holding disks to buffer some of the data from remote servers but most of the data to backup is already local. Now making the holdingdisk as large as the whole backup is ridiculous. This can't be right. Please provide a solution to this issue.

Ah, I made a mistake here. The DLEs that failed explicitly said "holdingdisk never". We did this because we did not want multiple parallel compressors (pigz) running which makes no real sense. For this I will find another solution or live with it. Sorry for the noise.

tanstafel commented 2 weeks ago

Just stumbled of the same issue. We upgraded from a CentOS 7 with amanda 3.3.3 to a Rocky9 with amanda 3.5.3 and our backup shows the same issue. We actually have a small holding disks to buffer some of the data from remote servers but most of the data to backup is already local. Now making the holdingdisk as large as the whole backup is ridiculous. This can't be right. Please provide a solution to this issue.

Ah, I made a mistake here. The DLEs that failed explicitly said "holdingdisk never". We did this because we did not want multiple parallel compressors (pigz) running which makes no real sense. For this I will find another solution or live with it. Sorry for the noise.

Spoke too soon: While removing the "holdingdisk never" is improved things - the large DLEs still do not work and complain with "out of holding space in degraded mode". So it looks like I really need a holding disk that fits the data of teh largest DLE. Most of this data is already present locally and it makes totally no sense that amanda copies this first to a holding disk and from there to the final destination. As I stated in my first post - this can't be right and there must be workable solution for this.