Re: [Duplicity-talk] duplicity 0.7 slowness

[Kenneth Loafman]

Just to verify, 

- Are both cache dirs actually in /var/.duply_cache?  

- Are both directories the same size (ls -l /var/.duply_cache)?  

- Same number of entries (ls -l /var/.duply_cache/* | wc -l)?

On Mon, Aug 1, 2016 at 1:07 PM, Philip Jocks via Duplicity-talk <address@hidden> wrote:

> Am 31.07.2016 um 22:52 schrieb edgar.soldin--- via Duplicity-talk <address@hidden>:
>
> On 31.07.2016 20:47, Philip Jocks via Duplicity-talk wrote:
>> Hej,
>>
>>> I am unable to reproduce this at all.  What is the filesystem where the archive-dir lives, what type?
>>>
>>> What OS?  Why not upgrade to 0.7.09 from the PPA or the tarball?
>>
>> It’s FreeBSD 10, archive-dir is UFS.
>> I wasn’t aware of 0.7.09 (duplicity.nongnu.org still links to 0.7.08 btw), just installed it with `python setup.py install —user` and seems to be way faster than 0.7.07.1 (13sec), which is good enough for me, though running it without —name is still <1sec.
>>
>
> weird, Phil could you please attach/post/pastebin the first 50 lines of a max verbose run '-v9'?
> please run once with -name and the other time without.

Here you go:

without —name: http://pastebin.com/raw/qCEKvm8D
with —name: http://pastebin.com/raw/hGREtEPk

This time, I noticed these lines when running with —name:

Processing local manifest /var/.duply-cache/duply_zzz_local/duplicity-inc.20160630T150316Z.to.20160630T150715Z.manifest (77341545)

[ PAUSE, where it is probably actually doing something ]

Found manifest volume 1
Found manifest volume 2
Found manifest volume 3
…
Found manifest volume 316
Found manifest volume 317
Found 317 volumes in manifest

So does —name cause duplicity to process manifest files that aren’t processed without —name?

Cheers,

Philip

_______________________________________________
Duplicity-talk mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/duplicity-talk