PasTim wrote: 
> I reduced the buffer size to 20MB, restarted LMS and did a full scan. 
> Not knowing quite how all this works, I then wondered if there might be
> relevant data held in memory even when LMS is stopped, so I rebooted and
> repeated.
> 
> To be be brief I have not detected much difference in scan times with
> changes in buffer size or with/without the extra indices.  Custom Scan
> is a few minutes quicker with a larger buffer, but it varies by several
> minutes each time anyway.  The LMS scan varies by a minute or so.
> 
> By far the biggest improvement for me in scanning has been due to the
> vacuuming.  I believe my Custom Scan scans are much quicker because they
> now start with a vacuumed library.db.  Similarly the improvement in menu
> responsiveness is, as far as I can tell, mostly down to vacuuming.  I'll
> try a test later without vacuuming.
I turned off all vacuuming, rebooted and checked auto-vacuum was still
off. Has something else changed in database tweak 0.6?  I've rescanned
since installing it, and library.db is always reported as having no
waste.  That wasn't the case when I first installed the previous
version.  The persist.db also seems to have low waste, but is ~30MB
bigger than it was.  Whilst custom scan scanning, the vacuum tick
appeared against persist.db and then disappeared. I'm a little puzzled.

The LMS scans currently take about the same as before, ~19 minutes.  The
Custom Scans are back up from 1.5 to nearly 2 hours.  Not the 3 hours
that I recall, and I don't know why.  It may be due to faulty memory
(mine, not the computer!), being on the latest ubuntu 14.04 (which I
only installed in mid April), or LMS 7.9, or some residual vacuuming
going on.

Even the first execution of my Custom Scan menus is not quite as
painfully slow as it was (~ 30 seconds).  

I wonder whether doing repeated rescans on an unchanged library is
causing less fragmentation than with a few changes or additions?

I still convinced by vacuuming, but am somewhat confused!  I had
expected different results, and I'm trying not to let my expectations
drive my interpretation of the the results.



LMS 7.9 on VortexBox Midi running Xubuntu 14.04, FLACs 16->24 bit,
44.1->192kbps. Wired Touch + EDO, coax to Musical Fidelity M1 CLiC.
Wireless Xubuntu 14.04 laptop controls LMS via Chromium.   Meridian
Explorer USB DAC to listen via Squeezelite on Vortexbox & other PCs as
required.  Spare Touch in loft.
------------------------------------------------------------------------
PasTim's Profile: http://forums.slimdevices.com/member.php?userid=41642
View this thread: http://forums.slimdevices.com/showthread.php?t=101469

_______________________________________________
beta mailing list
[email protected]
http://lists.slimdevices.com/mailman/listinfo/beta

Reply via email to