Hi Marcus,
My guess, from the "not in any space" from that one file you gave the
drain output for, is that all 40 of these files are not in any space
token and you've run out of "non-space token" space on your DPM.
How much space do you have in your pool when you run a "dpm-qryconf" on
your DPM headnode? (The space is free is listed in the first few lines
of the output).
It might be that you need to do some space token juggling to complete
this drain.
Hope that helps,
Matt
On 08/02/16 10:14, Marcus Ebert wrote:
> Hi,
>
> I try to drain one of our server (6 file systems, 9TB each) using
> dpm-drain which went fine mostly.
> For some files I get the error:
> "...is recorded as being in the process of being deleted, ignoring it
> during drain"
> which I think is nothing to worry about since it is marked to be
> deleted already anyway.
>
> However for about 40 files which are all on the same file system on
> this server, I get the following (example with one of the files):
>
> 02/08 08:58:49 8241,0: File:
> /dpm/ecdf.ed.ac.uk/home/atlas/atlaslocalgroupdisk/SAM/testfile-DEL-ATLASLOCALGROUPDISK-1453124939-8cd47c701ab3.txt
> 02/08 08:58:49 8241,0: pfn:
> pool5.glite.ecdf.ed.ac.uk:/mnt/gridstorage05/atlas/2016-01-18/testfile-DEL-ATLASLOCALGROUPDISK-1453124939-8cd47c701ab3.txt.88488079.0
> (of 1)
> 02/08 08:58:49 8241,0: replica type: primary
> 02/08 08:58:49 8241,0: file type: permanent (does not expire)
> 02/08 08:58:49 8241,0: size: 20
> 02/08 08:58:49 8241,0: space: not in any space
> 02/08 08:58:49 8241,0: replicating...
> dpm_getstatus_putreq: Failed for all SURLs
> 02/08 08:58:50 8241,0: failed
> 02/08 08:58:50 8241,0: dpm_replicatex
> pool5.glite.ecdf.ed.ac.uk:/mnt/gridstorage05/atlas/2016-01-18/testfile-DEL-ATLASLOCALGROUPDISK-1453124939-8cd47c701ab3.txt.88488079.0:
> No space left on device
>
>
> On all the other servers we have here, there is still enough space
> left. Why does it say "No space left on device"?
>
>
>
> Cheers,
> Marcus
>
|