From: Gary Thomas <gary@mlbassoc.com>
To: openembedded-devel@lists.openembedded.org
Subject: Why is OE so slow?
Date: Thu, 20 May 2010 07:02:18 -0600 [thread overview]
Message-ID: <4BF532DA.4080601@mlbassoc.com> (raw)
I'm using OpenEmbedded to build Angstrom. Recent trees
in the org.openembedded.dev branch have some 8370 recipes.
When I initiate a build, it takes *forever* just to get started.
I have [reasonably] powerful build machines (4 way 3GHz Zeon) and
it takes over 10 minutes CPU time just to reach 'Preparing runqueue'.
It then takes an additional 10 minutes before it starts to
execute the run queue. Of course, I have the Psyco JIT installed.
* What takes so long? Obviously, there are some processes in
these steps which are non-linear. I also build Poky which
has about 10% of the recipes (~850) and these steps are *much*
more than 10 times as fast. In fact, building Poky, the first
task in the run queue is usually started within 30 seconds...
Also, the "cache" (whatever it holds) does not affect these times
very much. The times above are for a first run (rebuild the cache).
If the cache does not have to be rebuilt, the first phase only takes
5.5 minutes (still very long), while the second phase (preparing
the run queue) still takes 10+ minutes (maybe longer now that it
has to consider partial results from the working build tree)
* Is there any way to tune/improve this? Is there something I've
missed? Why don't the online builders, e.g. narcissus, suffer
such delays?
I'd really like to use OE more, but having to wait 25 minutes
just to get it going is [at least] off-putting.
--
------------------------------------------------------------
Gary Thomas | Consulting for the
MLB Associates | Embedded world
------------------------------------------------------------
next reply other threads:[~2010-05-20 13:06 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-05-20 13:02 Gary Thomas [this message]
2010-05-20 13:28 ` Why is OE so slow? Andrea Adami
2010-05-20 13:53 ` Gary Thomas
2010-05-20 13:56 ` Frans Meulenbroeks
2010-05-20 14:27 ` Gary Thomas
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4BF532DA.4080601@mlbassoc.com \
--to=gary@mlbassoc.com \
--cc=openembedded-devel@lists.openembedded.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.