From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: Alex Kiernan <alex.kiernan@gmail.com>,
Patches and discussions about the oe-core layer
<openembedded-core@lists.openembedded.org>
Subject: Re: [OE-core] sstate cache management
Date: Wed, 22 Feb 2023 20:33:53 +0000 [thread overview]
Message-ID: <3983cc0aa3ab47e2b726103ab9dbd2a5408c76bb.camel@linuxfoundation.org> (raw)
In-Reply-To: <CAO5Uq5Qe4+kJANi+HuDrNK=9okEj9XgaFYzhhuixpVqdtZft1w@mail.gmail.com>
On Wed, 2023-02-22 at 17:56 +0000, Alex Kiernan wrote:
> I needed to do something about our shared sstate store and waded into
> the sstate cache management problem as the existing script takes hours
> to run over NFS (which for better or worse is where ours is). I've set
> myself the problem of replacing the existing script with something
> more extensible, understandable and performant.
>
> I've got something which I believed was roughly right, but I'm ending
> up with questions I can't answer when comparing the two outputs...
>
> If I run the existing shell script against a tiny sstate-cache (on my
> laptop) I get 420 duplicate files eligible for removal, if I run my
> script I get 491, looking into the delta, I pick out things like
> these:
>
> $ find sstate-cache/ -name
> sstate:libcap-ng:core2-64-poky-linux:0.8.3:r0:core2-64:10:*_populate_sysroot.tar.zst*
> -ls
> 49067 16 -rw-r--r-- 1 alexk alexk 14435 Feb 18
> 15:29 sstate-cache/25/59/sstate:libcap-ng:core2-64-poky-linux:0.8.3:r0:core2-64:10:2559429e2a553085bc657f3d2a21a111818061448e5fa2aa16398afb5dad8b90_populate_sysroot.tar.zst
> 49129 16 -rw-r--r-- 1 alexk alexk 15205 Feb 18
> 15:29 sstate-cache/25/59/sstate:libcap-ng:core2-64-poky-linux:0.8.3:r0:core2-64:10:2559429e2a553085bc657f3d2a21a111818061448e5fa2aa16398afb5dad8b90_populate_sysroot.tar.zst.siginfo
> 2490392 16 -rw-r--r-- 1 alexk alexk 15204 Feb 20
> 13:24 sstate-cache/bf/08/sstate:libcap-ng:core2-64-poky-linux:0.8.3:r0:core2-64:10:bf08f26e6bc5ab56ed128441fb05edeef41aa881330d04eea127a93ede51713d_populate_sysroot.tar.zst.siginfo
> 339439 16 -rw-r--r-- 1 alexk alexk 14423 Feb 20
> 13:24 sstate-cache/bf/08/sstate:libcap-ng:core2-64-poky-linux:0.8.3:r0:core2-64:10:bf08f26e6bc5ab56ed128441fb05edeef41aa881330d04eea127a93ede51713d_populate_sysroot.tar.zst
>
> Which look to me like I should be able to delete the older ones, or am
> I missing something? Trying to follow what the existing script is
> supposed to do is challenging!
I'd say delete the older one but it does depend a lot on what you're
building against the cache (e.g. multiple releases). The system is
meant to touch files it uses and the autobuilder just ages out things
not touched for X time where X has varied depending on the pressure on
our NAS.
Cheers,
Richard
next prev parent reply other threads:[~2023-02-22 20:34 UTC|newest]
Thread overview: 5+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-22 17:56 sstate cache management Alex Kiernan
2023-02-22 19:36 ` [OE-core] " Alexandre Belloni
2023-02-22 19:54 ` Alex Kiernan
2023-02-22 20:33 ` Richard Purdie [this message]
2023-02-22 21:12 ` Alex Kiernan
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=3983cc0aa3ab47e2b726103ab9dbd2a5408c76bb.camel@linuxfoundation.org \
--to=richard.purdie@linuxfoundation.org \
--cc=alex.kiernan@gmail.com \
--cc=openembedded-core@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.