eddy89 Apprentice


Joined: 01 Feb 2006 Posts: 180 Location: /world/Italy/Torino
|
Posted: Mon Jan 02, 2012 4:10 pm Post subject: Slow emerge -pv, database migration? |
|
|
Hello, I'm running gentoo on a quite slow machine (ARM 400MHz, 64Mb RAM).
I installed it in 2009, updated just as needed, but now I found that emerge is much faster (in newer stage3s).
I successfully tried to update python and portage, and I happily noticed that the default USE changed from berkdb to gdbm, I thought that it was the reason of performance increment, but it didn't. (btw I correctly updated python, python-updater, portage, emerge --sync etc.)
I read this message: Code: | * You are migrating from =dev-lang/python-2*[berkdb] to =dev-lang/python-2*[-berkdb].
* You might need to migrate your databases. | But I didn't found how to do so.
At the end, I tried to download a current stage3 for this machine, mounted --bind portage, chroot and tried the difference (emerge -pv transmisson):
On chrooted env. :
Code: | real 0m50.336s
user 0m47.350s
sys 0m1.150s |
On main root:
Code: | real 3m40.494s
user 3m28.280s
sys 0m1.760s |
Note that it's most "user" time, so cpu usage, not i/o (for those that would suggest defragmenting etc.)
What have I to do to achieve this performance on my main root? Has it something to do with "database migration"? Or some cache update? Python optimization?
P.S. emerge with no options takes ~15 seconds on both environments.
Thanks |
|